Connect · Example build
Ask your business: one assistant for your numbers and your knowledge
One assistant for your numbers and your knowledge. It queries your ERP or database, reads your documents, and shows where each answer came from.
Ask your business is a private AI assistant that answers questions from your own systems. Ask about numbers and it writes a read-only query against your ERP or database and runs it. Ask about policies, contracts or past decisions and it searches your documents and reads the relevant passages. Either way, it shows its work: the query it ran or the documents it cited.
It lives where your team already works, in Microsoft Teams or a simple web app, and it only shows each person what they're allowed to see.
The problem
Answers live in your ERP, your database and thousands of documents, so simple questions take a report request or a day of digging.
Signs you need it
- Simple questions ('What did we sell to this customer last quarter?') need a report request or a day of digging.
- The answers people need are split between an ERP, a database and thousands of documents.
- A few people know where everything is, and everyone else asks them.
- You've tried a chatbot over your documents and couldn't trust its answers.
How it works
- 1
Ask a question in plain English, in Teams or a web app.
- 2
For numbers, it writes a read-only query against your ERP or database and runs it.
- 3
For knowledge, it searches your documents, policies and email and reads the relevant passages.
- 4
It answers with its work shown: the query it ran or the documents it cited.
Try the demo
Runs on sample data in your browser. Nothing you do here is sent anywhere.
ASK YOUR BUSINESS · TEAMSREAD-ONLY · SALES ROLE
Pick a question. Numbers come from the database, policies come from your documents.
Work shown
The query it runs, or the documents it reads, appear here.
Simulated with sample data for a fictional distributor.
Example uses
Sales and customer questions
Top customers by revenue, open orders, last purchase dates and price history, answered from the ERP with the query shown.
Operations and inventory
Stock levels, late shipments, vendor lead times and margins, in plain English, without building a new report.
Policy and contract lookups
What a contract says about renewal, what the travel policy allows, how a past exception was handled, with the passage quoted and linked.
Onboarding
New hires ask the questions they'd otherwise ask a busy colleague, and get answers grounded in your own documents.
How we build it
Discovery starts with the questions your team actually asks. We collect them, agree on the definitions behind the important metrics (what counts as 'revenue', which date defines 'last quarter'), and map each to the tables or documents that answer it.
For numbers, the assistant generates SQL against a curated, documented view of your data, runs it read-only, and returns the result with the query attached. For knowledge, documents are indexed with their permissions, and answers cite the passages they came from (retrieval-augmented generation, or RAG).
Access follows your existing identity system, so the assistant can't show a person a document or a number they couldn't already open. Every question, query and answer is logged.
Where people stay in charge
Read-only by default. Every answer shows the query or document behind it, and it respects who's allowed to see what.
How it’s tested
Every build ships with an eval suite built from your real examples, re-run on every change, before and after launch.
We build a bank of real questions from your team with known-correct answers, then check every version of the assistant for the right numbers and the right sources, not just plausible-sounding text.
The question bank runs on every change. If a model upgrade gets a known answer wrong, it doesn't ship.
What you get
- An assistant in Teams or a web app
- Agreed definitions for the metrics that matter
- Permission-aware search over your documents
- A log of every question and query
- The source code and documentation, deployed in your environment. You own both.
We start with paid discovery: we learn how the work is done today, look at your real documents and data, and agree on what to build and what it has to get right. Then we design and build it, as a project with a not-to-exceed cap or on a Lab retainer. It runs on leading commercial models under terms that don't allow training on your data, or on a private model in your own environment.
Questions
Can an AI assistant query our ERP directly?
Yes, read-only. It writes a query against a documented view of your ERP or database, runs it, and shows you the query with the answer, so you can check exactly how the number was produced.
How do you stop it from making things up?
Numbers come from queries it actually runs, and knowledge answers cite the documents they came from. It's tested against a bank of your real questions with known answers, and it says so when it can't find an answer.
Will people see data they shouldn't?
No. It inherits permissions from your identity system, so each person only gets answers from data and documents they could already access.
Where does it run?
In Microsoft Teams or a web app, connected to your systems. The models can be commercial models under terms that don't allow training on your data, or a private model in your own environment.
Last updated October 20, 2026
Need something like this?
Tell us what you’re working with. We’ll reply with a few questions and a time to talk.
Related builds
AI connectors (MCP)
Gives ChatGPT, Claude or Copilot (AI tools you already pay for) secure access to your ERP, database and files, behind your single sign-on.ChatGPT, Claude, Copilot → your systemsConnects toEntra ID / Okta · ChatGPT · Claude · Copilot · ERPDocument ingestion AI
Reads invoices, contracts, statements and forms, pulls out every field, checks it, and writes it into your system.Stacks of PDFs → records in your systemConnects toQuickBooks · ERP · SharePoint · EmailPrivate AI
Open-weight models running in your own cloud or data center. Tuned only when testing on your real examples shows it helps.Your data → never leaves your serversConnects toAWS · Azure · Google Cloud · On-premise