The ontologyyour AI runs on.

Tacit models your company’s data, rules, processes and actions in one place, so every person and every AI agent works from the same reality.

See the platform Book a call
  1. 01The ontologyOne model of how your company works.
  2. 02BenefitsWhat your teams get: answers, agents and actions.
  3. 03Fits your systemsOn top of the systems you already run.
  4. 04How we workOur engineers on site, quarter by quarter.
01The ontology

A digital twinof your business.

Your ontology holds your data, rules and actions in one place, checked before anyone relies on them, so people and AI get the same answer.

PeopleAI agentsOntologyCustomerRefundOrderPolicyGDPRPaymentInvoiceTeamFinanceOperationsSalesAISupport agentAIRefund agentAIWeekly reportTacitFDE

What does an ontology capture?

  1. 01Data

    Connect every system once.

    ERP, CRM, documents and spreadsheets become one connected picture of your business.

  2. 02Logic

    Write the rules down.

    What counts as a late order or a good margin is defined once and checked before anyone relies on it.

  3. 03Actions

    Let AI act, with approval.

    Actions run through approvals and are written back to your systems, like SAP.

  4. 04Governance

    Govern people and agents alike.

    The same permissions apply to every person and every AI agent.

Running on Tacit today

591objects

41processes

68versions

38systems

02Benefits

Put your ontologyto work.

Every app runs on the same model of your company, so answers agree and actions follow your rules.

Ask in plain words and get the answer from your own data, with every step shown.

Book a call

See where your team’s time goes, find the repetitive work, and let AI take it over.

Your ontology gives every agent its context: data, rules, approvals and processes. Your developers build the rest in Claude, Cursor or Copilot.

The spec writes itself from what the ontology knows, then ships into your product.

They notice what needs attention and set the right actions going across your tools; a person approves what matters.

An ontology does more than power AI.

Once your company is written down, it pays off in places you might not expect.

  • Company brain

    What people know stays when they leave.

  • Fact check

    Every number traced to its definition and source before it reaches a board deck.

  • Compliance map

    Each regulation linked to the processes it governs.

  • Due diligence

    Hand investors or acquirers the model, not a round of interviews.

  • Onboarding

    New joiners read how the company really works, instead of asking around.

  • Audit prep

    Show auditors how every number was made.

  • Change impact

    See what breaks before you change a process.

  • Post-merger integration

    Map two companies onto one model.

03Fits your systems
Underneath
Your systems stay.Nothing to rip out: Tacit reads from SAP, Salesforce, HubSpot or Postgres and writes back.
On top
Any AI plugs in.Claude, ChatGPT, Copilot or your own agents, all through MCP.
Your data
Customer data stays put.The ontology knows what a customer is, never who your customers are.
Every change
Nothing changes without a person.Each edit is checked automatically and approved by someone; every version is kept.
04How we work

We come with engineers,not a login.

They work on site with your team, one business area at a time, until your people run it themselves.

Q1

Model one business area

  • Pick the area together
  • Build its ontology with your team
  • Connect its systems
Q2

Prove it

  • First agents and AI workflows
  • Built with your team
  • Tested on real work
Q3

Scale it

  • Roll out to more teams
  • Model the remaining areas
  • Train your people to work with AI
Q4

Go end to end

  • Agents that span systems
  • Whole processes, not single steps
  • Governance on every change
Q5

Let every team build

  • Teams build their own agents
  • One shared model underneath
  • Our engineers step back
Q6

Fully yours

  • Your team runs and extends it
  • We stay on call
  • The model keeps improving

We would rather be in the roomthan in your ticket queue.

Regular SaaS products

  • You buy seats.
  • You read the docs alone.
  • You set it up yourself.
  • Stuck? You file a ticket and wait.

Success meansPeople log in.

Tacit (SaaS + FDE)

  • An engineer sits with you from day one.
  • We learn your processes where they happen.
  • We build it with your teams, not for them.
  • Questions get answered in the room.

Success meansAn agent in production, run by your people.

Built by a boutique team in Berlin that has worked at

SiemensOnSAPZalando
05Next step

Does your AI know your business?Let’s create your first ontology.

Book a call.

Leave your work email. We’ll send you a link to pick a time, and you’ll see Tacit working live.

We use your email only to reply. Privacy policy

Why we’re called Tacit

tacit/ˈtæs.ɪt/ adjective

Known by doing, hard to put into words.

As the model fills in, your metrics turn from red to green.

  • The steps, in order.
  • Who signs off above €50.
  • When a refund is done.
  • Which customers to trust.
  • When to bend the rule.
  • Where money needs a second look.
  • Who is about to leave.
  • What’s really behind a spike.
  • Who may decide: person or agent.