For Leaders whose AI pilot works in a demo and nowhere else
AI systems
TechFabric builds the AI systems that sit between a lakehouse and an application: context stores, long-term memory, hybrid and contextual retrieval, semantic definitions, durable agent execution, evaluation harnesses, and permissions inherited from Unity Catalog. We run our own products on the same architecture, so the patterns we recommend are ones we operate.
Context stores, memory, retrieval and governed agents that survive production.
7 common questions, answered below ↓Agent demos run in a notebook and die on the first timeout. Getting one into production means solving durability, approvals, sandboxing, model routing, evaluation and cost accounting, and no framework tutorial covers those. We build the layer between the lakehouse and the application: the context store the agent reasons over, the memory that outlives the process, the semantic definitions that stop two answers both being right, and the governance that makes the whole thing explainable to a security team. We start from the scoring: a written rubric for what a good answer looks like, and an environment that can run it. A system nobody can score is a system nobody can improve.
- An eval rubric and a test environment before a line of the agent is written
- Context stores and memory that outlive the process
- Hybrid and contextual retrieval grounded in your own governed data
- Durable agent execution, so a long job survives a restart
- Regression suites that catch a bad answer before a board meeting does
How an engagement works
01
Talk to an engineer
A real conversation about your initiative with a senior engineer who has built this before. Not a sales call. What you are trying to build, what has been tried, and what is realistic.
02
Discovery and scoping
Two to three weeks to clarify requirements, evaluate where AI fits, and define realistic scope. On AI work this is also where success gets defined precisely enough to score, because a goal nobody can measure cannot be hillclimbed. You get a plan you can act on before committing to a larger engagement.
03
The right team, daily demos
We put the team the work actually needs on it and show you running software every day. Built with the same rigor as any enterprise system: tested, monitored, documented.
04
Production and beyond
Deployed and running under real load, handling real business processes. Ongoing support and team continuity for whatever comes next.
Databricks implementation
Lakehouse to live application. Migration, governance, and native Databricks apps.
Forward-deployed teams
Product, design and engineering people who sit inside your business, find the real problem, and ship it.
Platform & SaaS development
Full product delivery, from first architecture to a system running under load.
APIs & durable systems
Long-running operations that survive restarts, retries, and partial failure.
FAQ
AI systems, answered
What actually is an AI system, as opposed to an AI feature?
A feature calls a model. A system is everything that has to be true for that call to be trustworthy on the hundredth day: where the context comes from and how it is retrieved, what the agent is allowed to touch and under whose permissions, what happens when a step fails halfway, how you know an answer is still correct after someone changed a definition, and what it costs. The model is the part that is nearly free. The system is the part that decides whether it ships.
Do we need a knowledge graph, or is retrieval enough?
Usually the honest answer is that you need definitions before you need a graph. Most failures we see are not retrieval failures. They are two teams meaning different things by the same word, so both answers are correct and one of them is wrong to the person reading it. We start with the semantic layer, add structured relationships where questions genuinely span entities, and do not sell a graph to a problem that a definition would have fixed.
Can this run without our data leaving our environment?
Yes, and that is the default. Applications and agents run in your workspace under their own service principal, inheriting Unity Catalog permissions. Hosting by us is available where it suits, but nothing about the architecture requires your data to move for it to be reasoned over.
How do you prove an agent is working?
An evaluation harness with ground truth you own. We sit with the people who already know the right answer, write the questions they actually ask, record what the answer should be, and run that suite against the system on every change. Without it you are relying on whoever last used it not to have hit the broken case.
Have you actually put one into production?
We run our own company on one. GTM Brain is a production Databricks App that senses signals, judges accounts and drafts outreach under governed autonomy, with durable workflows and human approval gates. Fabric is a commercial product built on the same foundation, and across our clientele it has taken delivery work that needed a team of ten down to three.
Where does reinforcement learning fit?
Later than most people expect, and it is not usually the first problem. Optimisation of any kind, whether reinforcement learning on the model or an automated loop tuning the workflow around it, needs something to optimise against: a rubric that says what a good answer is and an environment that can score one. Almost every team we meet wants the optimisation and has neither. Build the rubric and the environment first, and you get most of the benefit from ordinary iteration; skip them and reinforcement learning is an expensive way to hillclimb a hill nobody has measured.
Who writes the rubric, you or us?
Both, in the same room, and that is the part we will not do at arm's length. Andrew Ripley, who runs product here, sits in those sessions for exactly this reason. The rubric encodes what your business considers a right answer, which is knowledge you have and we do not. Our job is to interrogate it until it is specific enough to score automatically, then build the environment that runs it. An engineer who has never sat with the people who own the question cannot write it, which is most of why we deploy forward.