<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>NearSync blog: Building NearSync</title>
    <link>https://nearsync.ai/blog/building/</link>
    <atom:link href="https://nearsync.ai/blog/building/rss.xml" rel="self" type="application/rss+xml" />
    <description>What we are building: product, engineering and releases.</description>
    <language>en</language>
    <lastBuildDate>Mon, 16 Nov 2026 05:00:53 GMT</lastBuildDate>
    <item>
      <title>When code gets cheap, judgement is the work</title>
      <link>https://nearsync.ai/blog/when-code-gets-cheap-judgement-is-the-work/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/when-code-gets-cheap-judgement-is-the-work/</guid>
      <pubDate>Mon, 26 Oct 2026 05:05:00 GMT</pubDate>
      <dc:creator>Ausaff Shakeel</dc:creator>
      <category>Building NearSync</category>
      <description>AI writes a lot of our code now, and it still gets things wrong. Why deciding what to build, and checking it, matters more than it did before.</description>
      <content:encoded><![CDATA[<p>There is a popular argument that AI has become a flawless senior engineer: the machines no longer make mistakes, and the only thing left for people to do is design. Half of that is right.</p>
<p>Here is what we believe, from building NearSync every day with AI in the loop. AI writes a large share of our code. It is fast, it is often good, and it is regularly wrong in ways that look right. That makes two things more important than they used to be: deciding what to build and how it should feel, and checking that what got built is true. Neither of those is going away. Both are now most of the job.</p>
<h2>What changes when code is cheap</h2>
<p>For most of the history of software, writing the code was the slow part. A feature took weeks because someone had to type it, test it and fix it. That cost forced a kind of discipline. You thought hard before you started, because starting was expensive.</p>
<p>That cost has dropped a long way. A screen that took a week can take an afternoon. A migration script, a report, an importer for a new file format: the first draft now arrives in minutes.</p>
<p>The obvious result is that you can build more. The less obvious result is that you can build the wrong thing faster than ever, and fill a product with features nobody asked for. When typing stops being the bottleneck, the bottleneck moves to the decisions around it. What should this do? Who is it for? What should it refuse to do? How should it feel on a long Tuesday afternoon?</p>
<p>Those questions do not get easier because the code got cheaper. They get more frequent.</p>
<h2>AI writes plausible code, and plausible is not correct</h2>
<p>AI does make mistakes, and it is worth being specific about which ones.</p>
<p>AI rarely makes the mistakes people expect, like a missing bracket. It makes the mistakes a confident new hire makes. It calls a function that sounds right but does not exist in this part of the code. It copies a list of values that already lives somewhere else, and the copy quietly drifts. It writes a check that passes in the build and fails the first time a real customer clicks the button.</p>
<p>None of that is a reason to stop using it. It is a reason to treat its output the way you would treat any colleague's: read it, test it, and never assume a green tick means it works.</p>
<h2>Checking is a system, not a step</h2>
<p>The more code gets written, the more there is to check, and the checks themselves are code that can break. A test run can stop running and still look like a test run. A build can skip a check without anyone deciding it should. So the checking has to be watched as carefully as the product: it needs its own tests, its own alarms, and someone who notices when it goes quiet.</p>
<h2>How we check the work now</h2>
<p>Here is what stands between a change and our customers today.</p>
<ul>
<li><strong>More than 850 files of automated tests.</strong> Over 400 cover the web app, around 100 cover the mobile app, and just over 300 cover the back-end services that generate documents, handle messages and run the AI. The web app and back-end tests run on every proposed change and every night.</li>
<li><strong>Tests that try to break the tests.</strong> The code that handles money, permissions and keeping each company's data separate gets mutation testing every night. A tool deliberately introduces small bugs into that code and checks that our tests notice. The score has a floor, and lowering it needs a reviewed reason.</li>
<li><strong>Proving things by doing them.</strong> Every change builds a fresh copy of the whole system from scratch and then actually performs the paths everything else depends on, rather than only checking that the right pieces exist. A check that only counts things can stay green while the thing it counts no longer works.</li>
<li><strong>One way into production.</strong> Production only accepts changes that have already been through our staging environment. A check enforces that, so it does not depend on anyone remembering.</li>
<li><strong>A person reads it.</strong> AI drafts, and an engineer reviews, runs it, and owns what ships.</li>
</ul>
<p>None of this is unusual on its own. What has changed is how much of it we lean on. When one engineer with AI can produce in a day what a team used to produce in a week, the checks have to scale with the output, or the mistakes do.</p>
<h2>The other half of the work is taste</h2>
<p>Checking tells you whether the code does what you said. It cannot tell you whether what you said was right. That part is design, in the widest sense: what the product does, what it leaves out, and how it treats the person using it.</p>
<p>Some of our clearest decisions have nothing to do with code at all:</p>
<ul>
<li><strong>No amber, anywhere.</strong> We do not use amber to signal stress or push a decision. People spend their working day on this screen, and it should not feel like a production line.</li>
<li><strong>Leave actually means leave.</strong> On approved leave, calls do not ring and notifications arrive silently. The system knows you are off.</li>
<li><strong>Breaks are part of the work.</strong> Vitality is one of the rings on the home screen, next to the rings for the work itself.</li>
</ul>
<p>An AI can implement any of those in an afternoon. It would not have suggested any of them. Each one came from someone deciding what kind of product this should be, and then holding to it when a shortcut was available.</p>
<p>The same goes for the harder product calls. NearSync puts sales, finance, people and payroll, projects, support and chat in one system for UAE businesses, with AI working in every team. Deciding how a deal, an invoice and a payslip should relate, and what a manager should see first in the morning, is not a coding problem. It is a judgement about how a business actually runs.</p>
<h2>Where that leaves us</h2>
<p>People often say product design is the final moat. We would put it more plainly. When code gets cheap, the scarce things are knowing what to build, knowing how it should feel, and knowing that it works. AI helps with all three. It replaces none of them.</p>
<p>So we use it heavily, and we check it heavily. We write down what a feature must do before anyone writes it. We review every change. We keep a test suite large enough that a quiet mistake has somewhere to get caught, and we watch the suite itself.</p>
<p>If you want to see what that approach produces, the values behind it are on our <a href="/about/">about page</a>, and the way we hire for judgement is on <a href="/careers/">careers</a>. Or <a href="/pricing/">see what NearSync AI includes</a> and try it with your own team.</p>
]]></content:encoded>
    </item>
    <item>
      <title>Seven NearSync features worth finding in your first week</title>
      <link>https://nearsync.ai/blog/nearsync-features-worth-finding-first-week/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/nearsync-features-worth-finding-first-week/</guid>
      <pubDate>Mon, 26 Oct 2026 05:04:00 GMT</pubDate>
      <dc:creator>Karol Pelc</dc:creator>
      <category>Building NearSync</category>
      <description>A keyboard shortcut that turns a sentence into a meeting, pins that never go stale, and WhatsApp commands with a PIN. Seven NearSync features worth knowing.</description>
      <content:encoded><![CDATA[<p>Most people learn a new system the same way. They find the screen for their own
job, learn it well, and stop exploring. A sales lead lives in the pipeline. A
finance manager lives in invoices. An HR manager lives in leave and payroll.</p>
<p>That is sensible, and it also means some of the most useful parts of NearSync
stay hidden for weeks. They are not tucked away on purpose. They simply sit
between departments, or behind a shortcut, or in a place you only visit when
something has already gone wrong.</p>
<p>This post walks through seven of them. Each one is in the product today, and
each one is described here by what it does for you, not by what it is called
on a settings screen.</p>
<h2>1. A morning screen that is never wrong</h2>
<p>The first screen in Workspace is the <a href="/solutions/workspace/cockpit/">cockpit</a>.
It is a single page built from six cards: a banner for anything happening right
now, what needs your action, today's schedule, team chat, tasks due, and what is
coming up.</p>
<p>The part people miss is the first card. Leave requests, invoices waiting on
approval, contracts waiting on a signature and expense claims usually live in
four different tools. Here they arrive in one list, whichever department they
came from, and you can open the request or the invoice straight from it.</p>
<p>Two details matter more than they look:</p>
<ul>
<li><strong>None of it is AI.</strong> The cockpit is put together from your own data: which
approvals are yours, which invoices are overdue, which meeting starts in ten
minutes. There is nothing to double-check, because nothing is guessed.</li>
<li><strong>Each card loads on its own.</strong> A slow card never holds up the rest, so the
screen is useful the moment you open it.</li>
</ul>
<p>It also only ever shows what you are already allowed to see. There is nothing
to set up, and nothing to leak.</p>
<h2>2. One shortcut that turns a sentence into a task or a meeting</h2>
<p>Press <strong>Cmd + ;</strong> (Ctrl + ; on Windows) anywhere in NearSync and a capture box
opens. Type what you need in plain words and it becomes a task, a meeting or a
reminder. Hold the shortcut down instead of tapping it and you can say it out
loud.</p>
<p>The capture box understands a short set of @-words, and the same set works
everywhere you can add something:</p>
<ul>
<li><strong>A person's name</strong> adds them to the meeting. A colleague and a client both
work.</li>
<li><strong>@huddle</strong> makes it an internal call inside NearSync.</li>
<li><strong>"on meet", "on teams" or "on zoom"</strong> picks the meeting provider.</li>
<li><strong>@team</strong> tells your department.</li>
<li><strong>@cockpit</strong> flags the item for triage, so it reaches the people who
sort urgent work.</li>
</ul>
<p>One decision behind this is worth knowing. Huddles are for your own team only.
If you add a client to a huddle, NearSync switches the meeting to Google Meet
and tells you why, rather than sending a client a link they cannot use.</p>
<p>A sentence that holds several things at once, like "call Sara at 4, then a 30
minute review with the team after", is split into separate items. You see the
plan as an editable preview first, and nothing is created until you confirm it.
Any attendee email that does not match a real person on your team or your
contacts is dropped, so a meeting never goes to an invented address.</p>
<h2>3. Pins that always show where the deal is today</h2>
<p>A bookmark in a browser is a snapshot. A <a href="/solutions/workspace/pins/">pin</a> in
NearSync is not.</p>
<p>Pin a deal while it is at Qualified, and the pin shows Won on the day the deal
is won, with nobody touching the pin. The same goes for a contact, a company, a
task, an invoice, a document, a meeting, a project or a colleague's employee
record. The pin keeps a link to the real thing and shows its current state every
time you look.</p>
<p>From the same bar you can also paste a web link or type a note. That is the
habit worth building: one collection per customer, holding the deal, the
article you found about them and your own note from the last call. Collections
can stay personal or be shared with a team, and you can view one as a compact
list or lay it out on a canvas.</p>
<p>Sharing a collection never gives anyone access to a record they could not
already open. Pinning changes what is in front of you, not who can see what.</p>
<h2>4. Saved views that follow you to any device</h2>
<p>The deals, contacts, companies, leads, products, services and bookings lists
all come with ready-made views. Two examples:</p>
<ul>
<li><strong>My Deals</strong> shows only deals you own, with a Stagnation column that counts
the days since each deal last changed. Sort by it and the deals going quiet
rise to the top.</li>
<li><strong>Awaiting first contact</strong> shows the contacts nobody has spoken to yet.</li>
</ul>
<p>You can make your own on top of these: choose the columns, set the order, give
it a name. Views you make are saved to your profile rather than to your
browser, so they are there on your laptop and on the office machine alike.</p>
<p>A small thing we learned the hard way: when we improve a ready-made view, your
own views stay exactly as you left them. Only the built-in ones are refreshed.</p>
<h2>5. WhatsApp commands for your own team</h2>
<p>Most people meet NearSync's WhatsApp as a shared inbox for customers. The other
half is for staff. From the WhatsApp they already have open, your team can send
<a href="/solutions/whatsapp/commands/">short commands</a> and get an answer in the same
thread:</p>
<ul>
<li><code>#in</code> and <code>#out</code> to clock in and out from a site</li>
<li><code>#leave</code> with dates to request leave on the way home</li>
<li><code>#task</code> to add a task</li>
<li><code>#paid INV-042</code> to mark an invoice paid</li>
<li><code>#status</code> to see what needs them today</li>
<li><code>#help</code> to see the full list</li>
</ul>
<p>A command is not a back door. Each one checks the same permissions the app
does, so somebody who cannot approve an expense at their desk cannot approve one
from their phone either, and what they did is recorded against their name. A
personal PIN opens each day's session, and the two commands that move money,
<code>#paid</code> and <code>#approved</code>, ask for the PIN again every time.</p>
<p>You can also build your own commands in Operations, with the same permission
settings, without anyone writing code.</p>
<h2>6. Meeting notes that become follow-up tasks</h2>
<p>When the meeting recorder joins a call from your calendar, Scribe writes the
transcript with every speaker named. Once the call ends it produces a recap of
what was said, what was decided and what people committed to, and emails it to
everyone on the invitation.</p>
<p>The part teams discover later is what happens next:</p>
<ul>
<li><strong>The meeting appears on the customer's timeline</strong>, on the contact and the
company, so whoever picks up the account next can see it.</li>
<li><strong>When the meeting is linked to a project</strong>, the commitments become follow-up
tasks. Each task names the meeting it came from, and where the owner's name
matches exactly one person on your team, the task is assigned to them. Where
the name is ambiguous, the task is left unassigned rather than handed to the
wrong person.</li>
</ul>
<p>Those tasks appear in your <a href="/solutions/workspace/plan/">plan</a> alongside everything
else you owe, sorted into today, upcoming, later and done. You do not need a
second list for meeting actions. See <a href="/solutions/meeting-intelligence/recording/">how recording works</a>
for the detail.</p>
<h2>7. Approvals that move up when nobody acts</h2>
<p>Every approval in NearSync can have a clock on it. A leave request or an expense
claim that nobody touches for the hours you set does not sit in one inbox
forever. It moves to the next person up the reporting line, or to the head of
the department, depending on how you set the rule.</p>
<p>This is set up once by whoever manages <a href="/solutions/human-resources/teams-and-access/">teams and access</a>,
and the rules are written against roles and reporting lines rather than named
people. That is why they keep working when someone leaves or changes title. The
approval tiers can also require a receipt above an amount you choose, so an
expense claim cannot reach an approver without one.</p>
<p>The person who raised the request does not have to chase anyone. That is the
whole point.</p>
<h2>Why these stay hidden</h2>
<p>Look back over the list and a pattern shows. Almost every feature here crosses a
line between two departments:</p>
<ul>
<li>The cockpit gathers HR, finance and sales approvals into one list.</li>
<li>A pin follows a deal through the sales team's work.</li>
<li>A WhatsApp command marks a finance invoice paid from a phone.</li>
<li>A meeting recap reaches the customer's timeline and somebody's task list.</li>
<li>An escalation follows the HR reporting line to move a finance claim.</li>
</ul>
<p>That is also why they are easy to miss. Nobody owns them. The sales lead does
not think of the cockpit as theirs, and the finance manager does not think of
WhatsApp as a finance tool.</p>
<p>It is also why they work. Because the deal, the invoice, the leave request and
the meeting all live in NearSync, a pin can show a deal's real stage, and a
meeting can create a task without anyone copying it across. These features are
not separate add-ons. They come from having every department in one place.</p>
<h2>Where to start</h2>
<p>If you already use NearSync, try three things this week:</p>
<ol>
<li>Open the cockpit first thing tomorrow, before your inbox.</li>
<li>Press Cmd + ; and type your next meeting as a sentence, with the person's
name in it.</li>
<li>Pin the three deals you check most, and see them update without you.</li>
</ol>
<p>If you are still evaluating, the fastest way to see all seven working together
is a walkthrough with your own examples. <a href="/book/">Book a demo</a> and bring a real
deal, a real leave request and a real meeting.</p>
]]></content:encoded>
    </item>
    <item>
      <title>How to migrate from HubSpot, Zoho or Pipedrive</title>
      <link>https://nearsync.ai/blog/migrate-from-hubspot-zoho-pipedrive/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/migrate-from-hubspot-zoho-pipedrive/</guid>
      <pubDate>Mon, 26 Oct 2026 05:03:00 GMT</pubDate>
      <dc:creator>Ausaff Shakeel</dc:creator>
      <category>Building NearSync</category>
      <description>Moving CRMs without losing history: what to export, what to clean first, what comes across directly, and how to run both systems side by side briefly.</description>
      <content:encoded><![CDATA[<p>Most CRM migrations do not fail on the day of the import. They fail in the week
before, when nobody decided which pipelines matter, who owns the records of the
people who left, or what to do with five years of dead leads. The import itself
takes minutes. The preparation is the work.</p>
<p>This guide walks through a move off HubSpot, Zoho CRM or Pipedrive in the order
that keeps your history intact. It uses NearSync AI as the destination, because
that is the system we know in detail, but most of the steps hold wherever you are
going.</p>
<h2>Step 1: export everything before you touch anything</h2>
<p>Before you connect anything, take a full export from your current CRM and keep it
somewhere safe. Even if the new system pulls your data directly, the export is
your copy of the state you started from. If a mapping goes wrong, you compare
against it.</p>
<p>Export at least:</p>
<ul>
<li>Contacts, companies (Zoho calls them accounts, Pipedrive calls them
organizations) and deals, each as its own file</li>
<li>Leads, if you use a separate leads list</li>
<li>Notes, emails, calls, meetings and tasks, if your plan lets you export them</li>
<li>Your list of users, with their email addresses</li>
<li>Your pipelines and the names of every stage</li>
</ul>
<p>Keep the activity history even if you never import it. In NearSync the direct
connectors bring across your records, not the timeline of calls and emails behind
them. An archive you can search when a customer asks "what did we agree last
March" is worth the ten minutes it takes.</p>
<h2>Step 2: clean before you move, not after</h2>
<p>Migration is the one clean chance to leave the clutter behind. Every duplicate
you bring across is a duplicate someone has to merge later.</p>
<ul>
<li><strong>Decide what not to bring.</strong> Leads that went cold two years ago, companies
with no contacts, contacts who left. Archive them in the export instead.</li>
<li><strong>Fix emails first.</strong> Email is the field most systems use to match a person.
A broken or duplicated email turns one customer into three.</li>
<li><strong>Check phone numbers.</strong> Many UAE customers reach you on WhatsApp and never
give an email. Make sure their numbers are complete.</li>
<li><strong>Merge the obvious duplicates.</strong> Sort companies by name and scan the list.
"Acme", "Acme LLC" and "ACME" side by side are hard to miss.</li>
<li><strong>Agree your pipelines.</strong> If you run six pipelines in HubSpot and only use
two, decide now which ones move.</li>
</ul>
<p>Then, in the new system, set up what your records need to arrive into: custom
fields for anything the old CRM held that the new one does not ship with, your
pipelines and stages, and your team. Owners can only be matched to people who
already exist.</p>
<p>One more step people skip: switch off any automation that fires when a contact
is created, in both systems. An import that creates five thousand contacts can
otherwise send five thousand welcome emails.</p>
<h2>Step 3: know what comes across directly, and what goes by file</h2>
<p>This is where vendors are vague, so here is exactly what NearSync brings in
today.</p>
<p><strong>Directly from the vendor.</strong> You sign in to the old CRM from the Data Hub
(Operations, Apps and Import) and choose what to bring. Nothing is copied until
you choose.</p>
<table>
<thead>
<tr>
<th>Source</th>
<th>What the direct connection brings</th>
</tr>
</thead>
<tbody><tr>
<td>HubSpot</td>
<td>Contacts, companies and deals</td>
</tr>
<tr>
<td>Zoho CRM</td>
<td>Contacts, accounts (as companies), deals and leads</td>
</tr>
<tr>
<td>Pipedrive</td>
<td>Persons (as contacts), organizations (as companies), deals and leads</td>
</tr>
</tbody></table>
<p>Across all three, the import also keeps:</p>
<ul>
<li><strong>The links between records.</strong> A contact stays attached to its company, and a
deal to its company and people, so long as you bring them in the same import.</li>
<li><strong>Owners.</strong> Each user in the old CRM is matched to a person on your team by
email. You match the rest by hand, and anyone left over, such as a former
employee, goes to one fallback owner you pick.</li>
<li><strong>Pipelines and stages.</strong> For each pipeline in the old system you choose to
copy it as a new pipeline, map it onto one you already have stage by stage, or
skip it and leave its deals behind.</li>
<li><strong>Original created dates</strong>, so a customer you won in 2021 still shows as one.</li>
<li><strong>Email opt-outs</strong> from HubSpot and Zoho, so people who unsubscribed stay
unsubscribed.</li>
</ul>
<p>Two details are specific to how we built it. Leads come across as leads and are
never turned into contacts or deals on the way in. And a contact with a phone
number but no email is kept, because phone-only and WhatsApp-only customers are
common here.</p>
<p>If your accounting sits in Zoho Books, it connects separately and brings your
customers and sent invoices into Finance. Zoho connections work with whichever
Zoho data centre your account lives in.</p>
<p><strong>By file.</strong> Anything the direct connection does not cover comes in as a CSV or
Excel file. Spreadsheet import covers Sales, People, Finance, Projects, Marketing,
Support and Web. Download the template for each record type first: it carries the
exact columns expected. Common column names from HubSpot, Zoho, Pipedrive and
Salesforce exports are already recognised, so most of the mapping arrives filled
in and you correct it rather than build it. Each template also has a Source
record ID column. Fill it with the ID from your old CRM.</p>
<h2>Step 4: run a pilot, then the full import</h2>
<p>For anything over a few thousand records, import a slice first. A hundred
companies and their contacts take an hour to check. A wrong mapping across forty
thousand takes a month to unpick.</p>
<p>Import in this order, every time:</p>
<ol>
<li>Companies, so contacts have somewhere to attach</li>
<li>Contacts, so deals have people</li>
<li>Deals, which can now link to both</li>
<li>Everything else</li>
</ol>
<p>A deal imported before its company arrives with no company, and importing the
company later does not repair the link. It is the most common migration mess
there is.</p>
<p>When the import finishes, NearSync shows what came across and what did not, and
the full record stays in the Imports tab of Data Management.</p>
<h2>Step 5: run both systems side by side, briefly</h2>
<p>Few teams can switch off the old CRM the same afternoon. A short overlap, one or
two weeks, is sensible. Longer than that, and you are reconciling two sources of
truth by hand.</p>
<p>Three rules keep the overlap safe:</p>
<ul>
<li><strong>Pick one system per kind of record.</strong> While the old CRM is still live, it is
where deals are edited. NearSync is where people learn the new pipeline.</li>
<li><strong>Re-run the import at cutover.</strong> Running the same direct import again updates
the records it brought in the first time instead of duplicating them, so
anything added or changed in the old CRM during the overlap comes across.</li>
<li><strong>Know what a re-run does.</strong> On that second run the old system's values win.
If your team edited an imported deal in NearSync during the overlap, the re-run
writes the old CRM's version back over it. The one exception is a lead you
already converted in NearSync: it stays converted.</li>
</ul>
<p>Then set the old CRM to read-only, keep your export, and cancel the plan when
the contract allows.</p>
<h2>Step 6: check before you build on it</h2>
<p>Do not build reports or switch automations back on until you have checked the
data.</p>
<ul>
<li><strong>Reconcile counts.</strong> Companies, contacts and deals, against the old system.</li>
<li><strong>Spot-check twenty records by hand.</strong> Counts can match while content is
wrong. A phone number in the email field passes every count.</li>
<li><strong>Work the cleanup queues.</strong> Data Management has guided queues for deals with
no company, contacts with no company, deals with no pipeline, stale deals,
duplicates and messy formatting. Fix a whole queue at once, and undo any fix
within 24 hours.</li>
<li><strong>Walk your pipeline.</strong> Are the deals in sensible stages, with the right
owners?</li>
</ul>
<p>Note that the undo window covers fixes, not the import itself. An import cannot
be undone as one action, which is why the pilot and your export matter.</p>
<h2>How long it takes</h2>
<p>The import runs in minutes. The preparation takes a few days for most teams, and
it is where the value is: a migration is the only time a CRM gets cleaned all at
once.</p>
<p>If you are comparing systems before you move, see how
<a href="/resources/compare/hubspot/">NearSync compares with HubSpot</a> and
<a href="/resources/compare/zoho-one/">with Zoho One</a>. The step-by-step guide to
<a href="/help/operations/migrating-from-another-crm/">migrating from another CRM</a> covers
each screen, and <a href="/help/operations/importing-spreadsheets/">importing spreadsheets</a>
covers the file route. When your data is in, the
<a href="/solutions/operations/data-management/">cleanup queues</a> are where you start.</p>
]]></content:encoded>
    </item>
    <item>
      <title>One set of articles for your help centre and your web chat</title>
      <link>https://nearsync.ai/blog/help-centre-web-chat-one-knowledge-base/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/help-centre-web-chat-one-knowledge-base/</guid>
      <pubDate>Mon, 26 Oct 2026 05:02:00 GMT</pubDate>
      <dc:creator>Karol Pelc</dc:creator>
      <category>Building NearSync</category>
      <description>The articles you publish in Web HQ are what your web chat AI answers visitors from. How that works in NearSync, what happens on its own, and what does not.</description>
      <content:encoded><![CDATA[<p>An older post on this blog described NearSync as a layer for keeping content in step across other software. That was wrong. NearSync AI is one system for running a business: sales, WhatsApp and web chat, finance, HR and payroll, operations, projects and your website, with AI working in each of them.</p>
<p>There is one part of that old post worth keeping, though, because it points at a real problem. Most companies write the same answer twice. Once as a help article, and again as the script, macro or chatbot flow that answers the same question in a chat window. The two drift apart, and the day they disagree is the day a customer is told the wrong thing.</p>
<p>This post explains how we avoided that in NearSync, precisely: what happens on its own, what needs a person, and the decisions that cost us something.</p>
<h2>Where the articles live</h2>
<p>Your website tools live in Web HQ. For most businesses that means:</p>
<ul>
<li><strong>Knowledge Base</strong>, where you write, edit and publish articles, grouped into categories</li>
<li><strong>Site Structure</strong>, for the pages of your site</li>
<li><strong>Legal Manager</strong>, for terms, privacy and the other documents a site needs</li>
<li><strong>Forms</strong>, <strong>Reputation</strong> for reviews, and a <strong>Web Overview</strong> of how the site is doing</li>
</ul>
<p>The web chat itself is set up in Operations, under Web Chat, and the conversations arrive in the Support inbox next to your email and WhatsApp. So the chat is run by the team that answers customers, and the articles are written wherever your team writes them. What connects the two is the article.</p>
<h2>Two switches decide who can read an article</h2>
<p>Every article in the Knowledge Base has two switches.</p>
<ol>
<li><strong>Published or draft.</strong> A draft is only ever seen by your team.</li>
<li><strong>Customer-facing.</strong> This one is off by default. Turn it on and the editor tells you what it means: the web chat can answer customers from this article.</li>
</ol>
<p>An article reaches customers only when both are on. A published article that is not customer-facing stays internal. It is still useful, because your team can read it, but no visitor sees it and the chat will never quote it.</p>
<p>There is also an internal article type that cannot be made customer-facing at all. The switch is disabled in the editor, and if anything tries to save one with the switch on, it is turned off before it is saved.</p>
<p>We made customer-facing an opt-in on purpose, and it has a cost. The most common support question about the Knowledge Base is "I published it and it is not there". The answer is almost always that the second switch is still off. We could make that go away by turning the switch on by default. We will not, because the failure in the other direction is an internal pricing note or a draft refund policy read out to a customer by an AI. An article that is missing from the chat is an annoyance. An article that should never have been in it is a real problem.</p>
<h2>What happens when you save</h2>
<p>When you save an article, NearSync prepares it for the chat straight away. You do not press a separate button or wait for a nightly job.</p>
<ul>
<li>The article is split into passages of about 2,000 characters, with about 200 characters of overlap so a sentence is never cut away from the one before it.</li>
<li>Each passage is turned into a form the AI can search by meaning, not just by matching words.</li>
<li>The new passages replace the old ones for that article. Edit a price in an article and the chat stops quoting the old price.</li>
</ul>
<p>If you switch customer-facing off, unpublish the article or delete it, saving removes its passages from what the chat can see. Taking something back is the same action as putting it out, which matters when the thing you are taking back is wrong.</p>
<p>Behind that, a background check runs every 15 minutes. It looks for any article that should be available to the chat but is not, prepares up to 30 of them per run, and clears away passages from articles that no longer qualify. The save is how articles normally arrive. The check exists so that a save that failed quietly does not leave a hole that nobody notices.</p>
<h2>A decision that came from an incident</h2>
<p>That 15-minute check used to run every 3 minutes. In August 2026 the service we use to prepare articles hit its usage limit, and the frequent check kept retrying articles that could not be prepared, tens of thousands of times a day. Worse, the old order of operations removed an article's existing passages before preparing new ones. When preparation failed, articles that had been working became empty.</p>
<p>We changed three things:</p>
<ul>
<li><strong>New passages first, then the swap.</strong> An article's existing passages are only replaced once the new ones are ready. If preparation fails completely, the old passages stay and the failure is reported.</li>
<li><strong>A budget and a cut-off.</strong> Each run prepares at most 30 articles and stops after two articles fail in a row, so an outage elsewhere costs almost nothing.</li>
<li><strong>Every 15 minutes, not every 3.</strong> The check is a safety net, not the main path, so it does not need to be fast.</li>
</ul>
<p>The cost is that an article whose save failed can take up to a quarter of an hour to reach the chat. We think that is the right trade. A chat that keeps answering from yesterday's article is better than one that answers from nothing.</p>
<h2>How the chat uses an article</h2>
<p>A visitor types a question into your web chat. Before any AI is involved, the message is checked against your chat intents. New organisations start with six:</p>
<ul>
<li><strong>Greeting, thanks and farewell</strong> get a short written reply, with no AI call.</li>
<li><strong>Book a demo</strong> shows a booking card that links to your public meeting page.</li>
<li><strong>Support</strong>, for messages like "something is broken" or "this isn't working", hands the chat to a person. The conversation is marked for your team and a support ticket is opened, so the request is tracked rather than lost in a chat log.</li>
<li><strong>Pricing</strong> goes to the AI, with your articles.</li>
</ul>
<p>Anything that does not match a pattern is classified by the AI against the same list. Questions that end up with the AI, including pricing, follow the same path:</p>
<ol>
<li>The question is matched by meaning against the passages from your customer-facing articles.</li>
<li>The five closest passages above a similarity cut-off are given to the AI.</li>
<li>The AI is told to answer in two to four sentences, to stay within those passages, to cite them, and not to invent details they do not contain.</li>
<li>If nothing relevant is found, it is told to say so plainly in one sentence and offer to connect the visitor with a person.</li>
</ol>
<p>The visitor can mark each AI reply as helpful or not. When a person on your team claims the chat from the inbox, the visitor is told someone has joined and the AI stops replying. A single chat also has a ceiling of 60 AI replies. After that, the AI hands over to your team and says so, rather than going quiet.</p>
<p>The AI that writes the reply is the model your organisation has chosen for it, running on NearSync's managed AI, so there are no keys to set up.</p>
<h2>A bug the article editor exposed</h2>
<p>One honest detail. Chat intents can narrow the search to articles tagged for that intent. The seeded pricing intent did exactly that: it only looked at articles tagged "pricing". But the article editor has no way to set that tag. So for every new organisation, pricing questions were matched against nothing, and the AI answered them with no articles at all.</p>
<p>We found it on our own site in October 2026 and fixed it on 24 October. The pricing intent now searches every customer-facing article. And whenever a narrowed search finds nothing, the chat searches everything once more before answering, so an article you wrote is not ignored because of a tag you could not set.</p>
<h2>The help centre runs on the same articles</h2>
<p>The public view of a knowledge base follows the same rule as the chat. An article can be read publicly only when it is published, customer-facing, and in a category that is public and not archived. That rule is enforced underneath the software, not in the page that shows the articles, so a mistake in a page cannot publish something private.</p>
<p>Our own help centre at <a href="/help/">nearsync.ai/help</a> is built exactly this way, from the NearSync Knowledge Base. At the last build it held 306 articles in 28 categories, and they are the same articles the chat on this site answers from. When we correct an article, the help centre is corrected at the next build and the chat as soon as the article is saved.</p>
<p>If you are writing your own, our article on <a href="/help/web/the-public-knowledge-base/">how to create a public knowledge base</a> covers the practical side: start from your five most common support questions, title each article as the question somebody would type, and fix it the same day the thing it describes changes.</p>
<h2>What this means in practice</h2>
<ul>
<li><strong>Write the answer once.</strong> The article your customer reads is the article your chat answers from.</li>
<li><strong>Correct it once.</strong> An edit changes what the chat says from the next question onwards.</li>
<li><strong>Keep internal notes internal.</strong> Customer-facing is off until you turn it on, article by article.</li>
<li><strong>Let the chat hand over.</strong> Support requests, unanswered questions and long chats go to your team in the Support inbox, with a ticket where it is a support request.</li>
</ul>
<p>Web chat comes with every plan. Having the AI answer visitors in web chat and WhatsApp is an optional add-on, an AI agent seat, priced per agent on the <a href="/pricing/">pricing page</a>.</p>
<p>To see how the pieces fit, read about <a href="/solutions/web-chat/">web chat</a> and the <a href="/solutions/wiki/">wiki and knowledge base</a>, or <a href="/book/">book a demo</a> and ask the chat on this site a question first. It is answering from the same articles this post describes.</p>
]]></content:encoded>
    </item>
    <item>
      <title>Point of Sale that keeps selling when the internet drops</title>
      <link>https://nearsync.ai/blog/point-of-sale-that-keeps-selling-when-the-internet-drops/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/point-of-sale-that-keeps-selling-when-the-internet-drops/</guid>
      <pubDate>Thu, 17 Sep 2026 07:00:00 GMT</pubDate>
      <dc:creator>Shafaf Bhat</dc:creator>
      <category>Building NearSync</category>
      <description>The counter in NearSync now takes sales with no connection, dates them correctly, and cannot double-charge when the network returns. Here is how.</description>
      <content:encoded><![CDATA[<p>A shop does not close because the router does. Every counter product in this
category takes cash sales with the network down, and until this month ours took
none, because a sale was one call to the server and there was nothing to catch it
when that call failed.</p>
<p>That is fixed. The till now sells offline, opens offline, and settles up when the
connection comes back without ever charging anyone twice. This post explains what
it does and the decisions behind it.</p>
<h2>What happens when the connection drops</h2>
<p>Nothing visible. You keep scanning, the total keeps adding up, you take the money
and the customer leaves with a document.</p>
<p>The sale is written to the device rather than sent. When the connection returns,
queued sales replay oldest first, so the order they happened in is the order they
land in. Nobody has to remember to do anything, and there is no separate "sync"
button to forget.</p>
<h2>Why the till still works after a reload</h2>
<p>Selling offline only helps if the page was already open. A tablet switched on
before the router, or a reload on a bad connection, used to meet the browser's
offline page.</p>
<p>The till now keeps its own application shell on the device. Page loads try the
network first and fall back to the last good copy. The code and styles it needs
are kept from the first time it opens with a signal, and the list of what to keep
comes from the build itself rather than being typed by hand, so it cannot drift
out of date as the product changes.</p>
<p>Only the application is cached. Your data, the queue of unsent sales and anything
that talks to the server are left alone deliberately, because a stale price or a
cached stock figure is worse than no figure at all.</p>
<h2>Why a sale cannot be charged twice</h2>
<p>This is the part that matters most, and it is the reason offline selling is harder
than it looks.</p>
<p>Two things can go wrong. A sale is sent twice because the till retried. Or a sale
reaches the server, the reply never arrives, and the till believes it failed.</p>
<p>Both are solved the same way. Every sale is given its own reference <strong>before</strong> the
first attempt, not after. That reference travels with it, and the counter already
had a rule that no two sales on the same counter can share one. So if the same
sale arrives again, the server returns the document the first one created rather
than creating a second sale. A retry is safe by construction rather than by
someone remembering to check.</p>
<h2>What time the sale happened</h2>
<p>A sale taken at half past four and sent at six o'clock is a half past four sale.</p>
<p>The order, the receipt and the invoice's supply date all carry the time the sale
was actually made on the device, not the time it was received. That time is held
inside the shift, so a device with a wrong clock cannot date a sale outside the
day it was taken.</p>
<p>This is not tidiness. A supply date is what a tax return is built from, and a sale
dated by when the network recovered would put takings in the wrong period.</p>
<h2>What the till knows with no connection</h2>
<p>Before it goes offline, the counter takes one snapshot of everything it is allowed
to sell at that counter, in that counter's currency, with every code each item can
be scanned by. Barcodes, internal codes, whatever is on the box.</p>
<p>So scanning works offline, prices are right, and an item that is not sold at that
counter is not sellable there by accident.</p>
<h2>The tax invoice is not optional</h2>
<p>In the UAE this is where most counter software quietly fails.</p>
<p>The authority's position on simplified tax invoices is explicit: offering to print
a tax invoice on request does not satisfy the requirement. Delivery is mandatory.
A till that generates a receipt number and hands over nothing is in breach, not
merely incomplete.</p>
<p>So the counter issues a real document every time. It renders on screen first and
prints second, because delivery is required and a printer is not. The seller on
that document is the legal entity that owns the counter, with its registered name,
address and tax registration, and the till already knows which entity it belongs
to without being asked.</p>
<p>Receipts render in Arabic as well as English, because a document a customer cannot
read is not much of a document.</p>
<h2>Money going back the other way</h2>
<p>A counter that only takes money is half a counter. The till gives change, handles
a return at the point of sale, and produces the credit document that goes with it,
so a refund is a recorded transaction rather than an adjustment somebody makes
later in the accounts.</p>
<h2>Who is buying</h2>
<p>A counter sale can now name its customer. That matters less for a bottle of water
and a great deal for anything a business buys, because the document a business
needs is different from the one a walk-in needs, and it turns on one fact: whether
the buyer holds a tax registration.</p>
<h2>What we have not done yet</h2>
<p>Commerce is marked Alpha in the product and the counter is part of that. It is in
daily use and still moving.</p>
<p>Card terminals are not integrated at the counter; cash and manually recorded
payment work today. The offline queue holds sales, not stock movements, so stock
figures catch up when the connection does rather than being reserved while you are
offline. And the decision about whether a document is a full or a simplified tax
invoice is answered from whether the buyer holds a registration, which means a
buyer recorded without one gets the simplified document.</p>
<p>If you run a counter and want to try it, we would rather hear what breaks than
have you find out quietly.</p>
]]></content:encoded>
    </item>
    <item>
      <title>Manufacturing in NearSync and why it matters</title>
      <link>https://nearsync.ai/blog/manufacturing-in-nearsync-and-why-it-matters/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/manufacturing-in-nearsync-and-why-it-matters/</guid>
      <pubDate>Mon, 07 Sep 2026 16:43:18 GMT</pubDate>
      <dc:creator>Shafaf Bhat</dc:creator>
      <category>Building NearSync</category>
      <description>Recipes, production runs, work in progress and scrap now live in Commerce HQ. What manufacturing does in NearSync, and why we built it next.</description>
      <content:encoded><![CDATA[<p>Manufacturing is live in Commerce HQ. You can write down how a product is made,
run production against it, and see the cost sitting on the factory floor while a
run is open.</p>
<p>This post explains what it does, how the pieces fit, and why we built it at this
point rather than later.</p>
<h2>What manufacturing in NearSync does</h2>
<p>Making things is two separate jobs, and the module is split the same way.</p>
<p><strong>Recipes</strong> record how a product is made. What one run consumes, what it yields,
and what the labour costs.</p>
<p><strong>Runs</strong> are the making. You open a run against a recipe, start it, and finish it.</p>
<p>Keeping them apart matters more than it sounds. A recipe is a template, and
opening a run takes a copy of it. So when you change a recipe next month, every
run you already completed keeps the costs it was actually built with. Your history
does not silently rewrite itself because someone corrected a quantity.</p>
<h2>How recipes work</h2>
<p>A recipe names an output and a yield. One run of this makes twelve of that.</p>
<p>Against it sit component lines. Each line names a variant, a quantity per run, and
an expected scrap percentage, because most processes lose a predictable amount of
material and pretending otherwise makes every run look like it overspent.</p>
<p>A recipe also carries a conversion cost, which is the labour and overhead of doing
the work. Without it, a manufactured item appears to cost only its parts, which
understates it and quietly overstates your margin.</p>
<h2>How a production run works</h2>
<p>A run is two events with a gap between them.</p>
<p><strong>Starting</strong> consumes the components. They leave stock at the moment work begins,
because that is when they are actually gone.</p>
<p><strong>Finishing</strong> produces the goods. The output enters stock, valued at what the run
consumed plus the labour.</p>
<p>Between those two events the value has to be somewhere. It is not components any
more and it is not finished goods yet.</p>
<h2>Why work in progress is tracked separately</h2>
<p>That gap is work in progress, and NearSync tracks it as its own figure.</p>
<p>If you have three runs open, the screen tells you how much value is sitting in
them. That is money you have spent, that is not on a shelf and not yet sellable.
Businesses that do not track this are usually surprised twice: once at month end
when stock is lower than expected, and again when finished goods arrive at a cost
nobody planned for.</p>
<p>The manufacturing screen leads with shortages for the same reason. The expensive
mistake in production is starting a run you cannot finish, because the components
are consumed the moment you start and the half-built output sits on a bench until
the missing part arrives. Showing what a shortage would cost before you start is
worth more than showing it after.</p>
<h2>What happens when a run yields less than planned</h2>
<p>You planned twelve and got ten. The two you lost do not disappear from the
accounts.</p>
<p>The cost of the run is unchanged, so the same total divides across the units you
actually got. Your ten units each cost more than they would have. That is what
really happened, and it is the number you need when you price the thing.</p>
<h2>Why scrap is separate from moving stock</h2>
<p>These get confused constantly, and they are different questions.</p>
<p><strong>Moving stock is a location question.</strong> Goods in, to a shelf, to a picking
location. The quantity is the same, the value is the same, and nothing posts to
your accounts.</p>
<p><strong>Scrapping is a value question.</strong> The units are destroyed on purpose. They are
gone, and their cost has to leave the balance sheet and land in the profit and
loss, because you paid for something you no longer have.</p>
<p>So scrap is its own operation with its own name and its own posting. It is not a
transfer to a bin that happens to be a dead end.</p>
<h2>How manufacturing connects to purchasing and stock</h2>
<p>Manufacturing is not a separate island in NearSync, which is the reason to have it
here rather than in a second system.</p>
<p>Components are the same variants you purchase. When you receive goods against a
purchase order, that receipt is matched against the order and the supplier bill, so
what you were charged is checked against what you ordered and what arrived.</p>
<p>Stock on hand is a consequence of the movements rather than a number someone
maintains. Receipts, transfers, runs, scrap and sales all post movements, and the
quantity you see is what those add up to. There is no separate stock figure that
can drift away from the ledger, because there is no separate figure.</p>
<p>Barcodes and serials run through the same records. A code is a row, and a scan is
evidence attached to a real movement rather than a lookup that leaves no trace.</p>
<h2>Why we built this for UAE businesses</h2>
<p>Most businesses making things here are not large enough to run a dedicated
manufacturing system, and the ones they can afford tend to be inventory tools with
production bolted to the side. So production gets tracked in spreadsheets next to
accounting software that knows nothing about it, and the cost of a manufactured
item is worked out by hand, occasionally.</p>
<p>That is the gap. Not the businesses running a full manufacturing suite, who are
served. The ones making real products, carrying real stock, and costing it in a
spreadsheet because the alternative was priced for someone larger.</p>
<p>Because manufacturing sits with purchasing, stock, invoicing and tax in one
system, a finished unit already knows what it cost to make, and an invoice for it
already knows how it should be taxed. That is the part a second system cannot give
you, whatever it costs.</p>
<h2>Commerce is in Alpha</h2>
<p>Worth saying plainly: Commerce HQ is marked Alpha in the product, and manufacturing
is part of it. It is in daily use and it is still moving. Screens changed
underneath people this month as counting, codes and reservations landed.</p>
<p>If you are running production and want to try it, we would rather hear what breaks
than have you find out quietly.</p>
]]></content:encoded>
    </item>
    <item>
      <title>The Case for Micro-CRMs: Why Per-Seat Billing is a Tax on Growth</title>
      <link>https://nearsync.ai/blog/the-case-for-micro-crms-why-per-seat-billing-is-a-tax-on-growth/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/the-case-for-micro-crms-why-per-seat-billing-is-a-tax-on-growth/</guid>
      <pubDate>Mon, 27 Jul 2026 06:57:30 GMT</pubDate>
      <dc:creator>Shafaf Bhat</dc:creator>
      <category>Building NearSync</category>
      <description>Stop penalizing your own success. Learn why bespoke, owned Micro-CRMs are replacing bloated SaaS giants for agile teams.</description>
      <content:encoded><![CDATA[<h2>The Success Tax</h2>
<p>Imagine you run a sales team. You hire 5 new reps. This should be a moment of celebration - your company is growing!
But instead of celebrating, you're calculating.</p>
<ul>
<li>Salesforce: +$1,500/month.</li>
<li>Outreach: +$750/month.</li>
<li>Slack: +$80/month.</li>
</ul>
<p>Before these new hires have sold a single dollar, they have increased your burn rate significantly. This is the <strong>Per-Seat Business Model</strong>, and it is designed to tax your growth.</p>
<h2>Paying for Obesity</h2>
<p>Legacy CRMs (Customer Relationship Managers) like HubSpot and Salesforce are incredible feats of engineering. They are built to serve everyone from a 2-person startup to Coca-Cola.
To do this, they include <em>everything</em>.</p>
<ul>
<li>Territory Management? Check.</li>
<li>CPQ (Configure, Price, Quote)? Check.</li>
<li>AI Forecasting? Check.</li>
</ul>
<p><strong>The Problem:</strong> Your specific business likely uses only <strong>5-10%</strong> of these features. Yet, you pay for 100% of the code, 100% of the server load, and 100% of the complexity. You are effectively subsidizing the features used by Fortune 500 companies.</p>
<h2>Enter the Micro-CRM</h2>
<p>A <strong>Micro-CRM</strong> is the antithesis of the monolith. It is a lean, purpose-built application engineered specifically for <em>your</em> workflow.</p>
<ol>
<li><strong>Zero Bloat:</strong> If you don't need "Territory Management," it doesn't exist in the code. The interface is clean, fast, and requires zero training.</li>
<li><strong>SQL Ownership:</strong> Instead of your data being locked in a proprietary cloud, it lives in a standard PostgreSQL database that you own. You can connect it to Metabase, PowerBI, or any other tool directly.</li>
<li><strong>Unlimited Seats:</strong> Because you own the software, adding a user costs you $0. The only cost is the marginal server load, which is negligible (pennies).</li>
</ol>
<h2>The Economics of Ownership</h2>
<p>Let's look at the 3-year math for a team of 20:</p>
<p><strong>SaaS Rental (HubSpot Pro):</strong></p>
<ul>
<li>$100/user/month * 20 users = $2,000/month.</li>
<li><strong>3-Year Total:</strong> $72,000 (rented, zero equity).</li>
</ul>
<p><strong>Micro-CRM Build (NearSync):</strong></p>
<ul>
<li>One-time Build Fee: ~$15,000.</li>
<li>Hosting: $50/month.</li>
<li><strong>3-Year Total:</strong> ~$16,800 (owned asset).</li>
</ul>
<p><strong>Savings:</strong> Over $55,000.
More importantly, you possess a software asset that increases your company's valuation, rather than a liability on your P&amp;L.</p>
<p>Stop renting your customer relationships. Build the system that fits your hand.
    </p>
]]></content:encoded>
    </item>
    <item>
      <title>The &quot;Plugin Tax&quot;: The Hidden Cost of Generic Platforms</title>
      <link>https://nearsync.ai/blog/the-plugin-tax-the-hidden-cost-of-generic-platforms/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/the-plugin-tax-the-hidden-cost-of-generic-platforms/</guid>
      <pubDate>Sat, 18 Jul 2026 06:57:30 GMT</pubDate>
      <dc:creator>Tech Lead</dc:creator>
      <category>Building NearSync</category>
      <description>Why installing 25 plugins to get basic functionality is slowing down your site and draining your wallet.</description>
      <content:encoded><![CDATA[<h2>"There's an app for that!"</h2>
<p>WordPress and Shopify ecosystems love plugins. It sounds great at first. - Need a wishlist? Install an app ($10/mo). - Need reviews? Install an app ($50/mo). - Need a loyalty program? Install an app ($200/mo).</p>
<h2>The Reality</h2>
<p>Before you know it, your "cheap" $29/mo store is costing you <strong>$1,500/mo</strong>. </p>
<p><strong>Performance Cost:</strong>
These plugins are written by different developers. They load different scripts (jQuery, React, Vue) on your frontend. This bloats your site code, slows down loading times, and hurts your SEO.</p>
<p><strong>Native Architecture:</strong>
When we build a Custom OS, we code the Wishlist, Reviews, and Loyalty logic <strong>directly into your core database</strong>. - <strong>Cost:</strong> $0/mo extra. - <strong>Speed:</strong> Instant. - <strong>Conflicts:</strong> Zero.
    </p>
]]></content:encoded>
    </item>
    <item>
      <title>The Magic of APIs: A Guide for Non-Technical Founders</title>
      <link>https://nearsync.ai/blog/the-magic-of-apis-a-guide-for-non-technical-founders/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/the-magic-of-apis-a-guide-for-non-technical-founders/</guid>
      <pubDate>Fri, 17 Jul 2026 06:57:30 GMT</pubDate>
      <dc:creator>Education Team</dc:creator>
      <category>Building NearSync</category>
      <description>APIs are the digital glue of the modern economy. Here is a simple explanation of how they connect your world without code.</description>
      <content:encoded><![CDATA[<h2>The Waiter Analogy</h2>
<p>Think of an <strong>API (Application Programming Interface)</strong> like a waiter in a restaurant.</p>
<ol>
<li><strong>You</strong> are the Customer (The Frontend App).</li>
<li><strong>The Kitchen</strong> is the System/Database (The Backend).</li>
<li><strong>The Menu</strong> is the documentation (What you can ask for).</li>
</ol>
<p>You don't go into the kitchen and cook the burger yourself. You tell the waiter what you want ("Get me Order #123"). The waiter takes your message to the kitchen, the kitchen prepares the data, and the waiter brings it back to you.</p>
<h2>Why It Matters</h2>
<p>The waiter is the API. It is the messenger.</p>
<p>In your business, APIs let different systems talk: - Your <strong>Website</strong> asks the <strong>Stripe API</strong> to charge a card. - Your <strong>Inventory System</strong> asks the <strong>Shipping API</strong> to print a label. - Your <strong>CRM</strong> asks the <strong>WhatsApp API</strong> to send a message.</p>
<p>We build <strong>"API-First"</strong> platforms. This means your data isn't locked in a box. It has "handles" (endpoints) that let any other modern software connect to it securely. This is how we automate your manual work.
    </p>
]]></content:encoded>
    </item>
    <item>
      <title>The Agency Trap: Why Boutique Studios Need a Custom Operating System</title>
      <link>https://nearsync.ai/blog/the-agency-trap-why-boutique-studios-need-a-custom-operating-system/</link>
      <guid isPermaLink="true">https://nearsync.ai/blog/the-agency-trap-why-boutique-studios-need-a-custom-operating-system/</guid>
      <pubDate>Thu, 16 Jul 2026 06:57:30 GMT</pubDate>
      <dc:creator>Strategy Team</dc:creator>
      <category>Building NearSync</category>
      <description>Service businesses are notoriously hard to scale. Learn how building your own internal OS transforms you from a &quot;Time-Seller&quot; to a &quot;Productized Powerhouse&quot;.</description>
      <content:encoded><![CDATA[<h2>The "Feast or Famine" Cycle</h2>
<p>Running a boutique agency (Marketing, Design, or Consulting) is a constant battle. You win a client, you drown in work, you stop selling, the project ends, and you panic.</p>
<p>The culprit? <strong>Admin Overhead.</strong>
You spend 40% of your time chasing invoices, finding files in Google Drive, updating Trello cards, and copying data between Slack and Excel.</p>
<h2>The Productization Solution</h2>
<p>The most profitable agencies don't just sell "hours"; they sell a <strong>System</strong>.
Instead of cobbling together 5 SaaS tools, imagine a single <strong>Agency OS</strong>:</p>
<ol>
<li><strong>Client Portal:</strong> Clients log in to approve designs, pay invoices, and see progress. No more email threads.</li>
<li><strong>Asset Vault:</strong> All deliverables are auto-saved to a structured database.</li>
<li><strong>Auto-Billing:</strong> Milestones trigger invoices automatically.</li>
</ol>
<h2>Why Custom Wins for Agencies - <strong>Brand Perception:</strong> A custom portal with your logo makes you look like a tech-enabled enterprise, justifying higher fees. - <strong>Efficiency:</strong> You can handle 3x the clients with the same team because the admin work is automated. - <strong>Asset Value:</strong> When you want to exit, selling an agency with a proprietary software platform is worth 5x more than selling a client list.</h2>
<p>Stop renting your operations. Build the machine that runs your agency.
    </p>
]]></content:encoded>
    </item>
  </channel>
</rss>
