EpistemicWorks · knowledge engineering for AI agents
Knowledge engineering for AI agents

Your AI agent is only as good as what your experts can't say.

We specify, build, and run AI agents for SMEs. Every business has a way of doing things that actually works — pricing rules, qualification criteria, policies that took years to get right. We document it, specify it, and build it into your system, so your agent quotes correctly, qualifies leads properly, and handles policy questions reliably, even when your best people aren't in the room.

The problem

Your business runs on knowledge that nobody has written down.

Think about what your best people actually know — pricing rules that aren't in the rate card, clients who need handling differently, policy questions that have a right answer if you know who to ask. That knowledge is real, it took years to build, and it exists almost entirely in people's heads. Which means it disappears whenever they're not available, leave the business, or simply can't be reached.

What happens without it

An AI tool gets deployed with the best intentions but no real grounding in how your business works. It handles the obvious cases. It fails on the edge cases — the ones that require the knowledge nobody ever wrote down.

Every failure gets patched. The system accumulates fixes, but never a real understanding of your domain.

What we do instead

We make the knowledge explicit before anything is built. We interview your experts, document the rules and the exceptions, and specify it precisely enough to build from. Then we build the agent.

The result isn't a generic tool you've configured — it's a system that actually knows how your business works.

What we do

Three ways to work with us.

From thinking to building to running. Each tier includes everything in the one before it.

Tier 1
Specification Only
For teams that have developers but need the domain knowledge captured properly before anyone writes a line of code.
  • Structured interviews with your domain experts — remote or on-site.
  • A full behaviour specification — happy path, edge cases, failure modes.
  • Behavioural rules tied to real reasons — every rule linked to a specific risk, regulation, or past incident. Nothing in the spec is unexplained.
  • A runnable system prompt — built directly from the specification.
  • Line-by-line documentation — every fragment traceable to its source.
  • An evaluation set — concrete examples drawn from your operating reality.
  • Synthetic personas — adversarial and realistic users for stress-testing before go-live.
  • One handover session — with whoever will build or audit the system.
  • Workflow scoping included — we identify which single workflow to specify first for maximum ROI, so you're not paying to document everything before anything works.
Fixed price. Delivered in 2–3 weeks.
Get in touch to discuss →
Tier 2
Specification + Build
For owners who want the whole thing designed, built, and handed over — ready to run.
  • Everything in Tier 1, plus:
  • Full implementation — system prompt, integrations, tooling.
  • A basic evaluation harness — so you can keep testing as the domain evolves.
  • Handover documentation — written for the developers who will operate and extend the system.
  • One month of post-launch support.
Fixed price agreed at scoping.
Get in touch to discuss →
Tier 3
Specification + Build + Managed
For businesses that want AI working in the background without hiring anyone to maintain it.
  • Everything in Tier 2, plus:
  • We host and operate the agent.
  • Monitoring and incident response.
  • Prompt and evaluation updates — as your domain evolves or models change.
  • Monthly written summary — what the system did, what it flagged, what we changed.
Monthly retainer after initial build.
Get in touch to discuss →
Entry point
Specification Diagnostic
Not sure where to start? Not sure if your planned deployment has a specification gap?

We'll assess your planned AI deployment against our failure-mode framework and deliver a written readiness report: what's missing from your current spec, which edge cases are unaddressed, and whether a full engagement is warranted.

Fixed price. Delivered in 5 working days.

This is the fastest way to find out whether we can help — and the fastest way for us to find out whether you're a fit.

Get in touch to discuss →
"The knowledge that makes your best people good at their jobs shouldn't disappear when they're not in the room."
Fit

Who this is for.

We work with businesses that run on hard-won expertise — and where that expertise not being available costs real money.

A strong fit

  • SMEs with quote-to-cash workflows: manufacturing, trade, field services, professional services
  • Businesses where pricing rules, exceptions, and edge cases are known by two or three people — and took years to develop
  • Customer-facing workflows where a wrong answer costs a sale, triggers a return, or damages a relationship
  • HR and policy workflows where the right answer exists but employees can only get it by interrupting someone who knows
  • Owners who want to deploy AI without hiring an AI team — and who want it done properly the first time

Not a fit

  • Teams looking for the cheapest possible implementation
  • Engagements where the domain experts aren't willing to be involved
  • Consumer-grade chatbots with no compliance surface
  • Pure prototyping where speed matters more than rigour
Writing

How we think about the problem.

The methodology is documented. Read it before you hire us and challenge it when you do.

Start a conversation

If your business runs on expertise that took years to build, and you're thinking about AI, we should talk first.

A thirty-minute conversation. No deck. We'll ask what you're trying to automate, where the knowledge currently lives, and whether a specification engagement makes sense. Write to us at hello@epistemicworks.com — we reply within one working day.