AI in every department, under your control.
It works on the same deals, invoices and conversations your team already uses, running on the models you choose. Build your own agents, keep conversations you can come back to, and see exactly where AI is used and where it is not.
Every place it shows up
Where we use AI.
This is not one assistant bolted on the side. AI is built into the work itself, in every department, on the same deals, invoices and conversations your team already works with. Here is the full list.
- Sales ManagerReads your live pipeline, then answers and acts on it
- Deal briefA plain read on any deal, from its stage, age and last activity
- Commission workingA won deal produces the commission, already calculated
- Marketing ManagerReads both boards and the hours of work in each stage
- CreativeGenerates on brand, because the brand kit is an input
- Asset taggingSuggests tags for a file you just uploaded
- Campaign analysisReads real spend and channel mix, and writes the read
Comms
Workspace
- ConversationsA thread you come back to, with voice, on whichever model
- CoachHow you ran the meeting, and whether you are improving
- Meeting write-upSummary, decisions and tasks, emailed to everyone who was there
- The morning briefOne thing worth your attention, from your own records
- The StudioDescribe the chart you want and it builds it
- RecommendationsWhat to add next, validated against your real catalogue
- Empty charts explain themselvesWhich of four reasons applies, said in a sentence
- Metric draftingDescribe a measure and it drafts one, clamped to the catalogue
- WorkflowsDescribe an automation and it builds it as a draft
- A step that thinksPut the AI inside a workflow and branch on what it says
- Smart linksMatches a party-less deal to a company by meaning
- WhatsApp commandsPulls the fields out of a message people send in
Web
- Knowledge base authoringTidy, tag, rate, or draft a whole article unpublished
- Content polishGrammar and tone on something you wrote
Everywhere
Agents
Build your own agents.
Seven ready-made agents to start from. Point one at any part of your business, from deals to invoices to leave, give it a schedule, tick what it is allowed to do, and cap how much it can do in one run. No developer needed.
Finds deals that have gone quiet, after checking the tasks and the calendar
Watches invoices against the deals they came from
Turns what a meeting agreed into tasks with owners
Flags conversations nobody has answered
Finds records that are missing their links
Warns you when approved leave means nobody is covering
Sets a priority and category on new support tickets
One agent can look across several parts of your business at once. That is why Deal Reviver checks the calendar and the task list before telling you a deal has gone quiet.How agents work
Conversations
Conversations you can come back to.
Ask about your own business and keep the thread. Pin the ones that matter, pick them up next week, and talk to it out loud instead of typing. It runs on whichever model that department is set to, and it only ever sees what you can see.
Meetings
Fifty hours of meetings, included.
Transcription, the write-up and the coaching all run on real AI, and none of it touches your question credits. Fifty hours come with the plan.
Being straight about it
What we do not call AI.
Some of the most useful features in NearSync look like AI but run on plain rules over your own data, which is exactly why they are always accurate. They are good features. They are not AI, and we would rather tell you.
Setup
A different model for every department.
Route Engineering to one model and Finance to another, and change it later without touching anything else. Only an admin can, and the choice is visible rather than buried in a settings page nobody opens.
- SalesGPT-5.4Fast, and it talks to people all day
- FinanceClaude Opus 4.8Reasoning, where being wrong costs
- MarketingClaude Sonnet 4.6Writing
- OperationsGPT-5.4Builds your workflows
- EngineeringClaude Opus 4.8Long context
- CommsGemini 3.5 FlashHigh volume, every message
- SupportLlama, on your own machinesNothing leaves the building
- PeopleGemini 3.5 FlashEveryday questions
- WebGemini 3.5 FlashContent and tidy-ups
- AnalyticsGPT-5.4Turns a sentence into a chart
- NearSync AIMasterInherited by every HQ on Master
Change one and it applies to every session on the next prompt. A department left on Master follows whatever the base model is.
Alongside you
Point it at something and it drafts.
The reply, the summary, the follow-up, written against the record in front of you. Nothing it writes goes anywhere until you say so.
هل ما زلنا على موعدنا يوم الخميس؟
Translation
"Are we still on for Thursday?"
Shown to you only. The message itself is untouched.
Draft a reply
Yes, Thursday still works. I will send the agenda across on Wednesday.
It lands in your composer. You still press send.
Without you
The four things that act on their own.
Everything else waits to be asked. These four are the exceptions, and each one shows what it did and why rather than acting quietly.
What it does on its own
- Answers an inbound message from what is in your knowledge base
- Cites the article the answer came from
- Books nothing it cannot confirm, and sends your booking link instead
- Hands over to a person and opens a ticket when it should
What stops it
- Off until you turn it onAnd it cannot be turned on until you have said where escalations go.
- It defers when unsure"I do not have a confident answer to that, let me connect you with someone."
- A person takes over and it stopsThe moment somebody claims the conversation, it stands down.
- It cannot inventNo prices, discounts, dates or commitments that are not in your own articles.
This is the only one that talks to your customers. It is also the most heavily fenced, and it ships switched off.
Management
Credits, usage, and where they went.
What your team is actually using, per seat and per department, so a heavy month is something you watch arriving rather than read about afterwards.
- GPT-5.448%
- Claude Opus 4.832%
- Gemini 3.5 Flash13%
- Claude Sonnet 4.67%
Read on
Every capability, in depth.
Using it
The guides, in full
Plain guides for the person who sets it up, and for the person responsible for it.
Overview
Three layers, and where the AI turns up in your day.
Setting it up
Where the intelligence comes from, and which model answers.
Making it yours
Your material, your voice, and the limits it works inside.
Working without being asked
The scheduled AI that notices what nobody did.
Getting more from it
Good questions kept, and models tested against each other.
Watching it
What it costs, how fast it is, and whether it is working.
The questions we get asked.
Anything longer lives in the help centre.
Which AI models can we use?
Eight providers, or your own local Ollama. On Self-hosted you bring your own keys and there is no cap on what you spend.
Can different departments use different models?
Yes. Route Engineering to one model and Finance to another, and change it later without touching anything else.
How much AI do we get?
Two thousand credits a month per seat on Standard and four thousand on Pro. One credit is roughly one question, and you can top up.
Can we give it our own name?
Yes. Name it once and it is called that everywhere, with your logo, your colours and the guardrails you set for it.
Give every department its own model, and see it work on your own business.
Book half an hour and watch it answer questions about your own material, on whichever model you choose.