We already own the engine.
Custom regulated AI used to take 24 months. We ship in 6 to 12. The reason is the work that does not need to be redone for each engagement. The language the agents run on. The libraries the system is built from. The reference architectures the buyer's function is customised onto.
Same shape. New substrate.
Teklabs.iGentic and Teklabs.Analytics work across unstructured data the way SQL and a database engine work across structured data. SQL is the language; the engine executes it; together they let an organisation ask questions of structured data at scale, reliably, with audit. iGentic is the language; Analytics is the engine; together they let an organisation perform judgement-grade work across unstructured data at scale, reliably, with audit. Same shape. New substrate.
The work always exceeded the staff's attention. It required reading every record, every file, every artefact, against the standard. Structured data tooling could not do it; the records are largely unstructured and the work requires contextual judgement at the point of comparison. Reliable agents alone do not do it either; agents have to reach into the data with the spec intact. iGentic and Analytics together do.
Teklabs.iGentic, the language.
Teklabs.iGentic is our proprietary process definition language for AI agents. A structured plain-text grammar with named sections for role, intent, process, tools, anti-patterns, required behaviours, output, and audit. The agent runs iGentic as its working program.
Teklabs.Analytics, the AI engine on your data.
Teklabs.Analytics connects iGentic-specified agents to the client's data estate. Unstructured clinical records, document corpora, operational systems, the artefacts the work has to read against the standard. The agent scans the data, surfaces what matters, cites the source, and presents the result for staff action.
Three further libraries do the foundational engineering.
Three further libraries do the foundational engineering that lets the engine ship into regulated production.
Teklabs.Semantic is production-grade semantic search for .NET. It powers retrieval inside agentic workflows.
Teklabs.Security is multi-tenant identity, role-based access control, location-aware authorisation, and the cryptographic audit chain.
Teklabs.UI is the React component library the surfaces are built from. Teklabs.Kanban is the workflow lifecycle and task management substrate.
The libraries are not sold standalone. They are the engineering that lets us ship in months. Clients receive licensed access as part of a Teklabs delivery engagement.
Two architectures sit on top of the library set.
Two architectures sit on top of the library set.
What this means for the buyer.
Each engagement starts at the customisation layer, not the architecture layer.
The build cost is in proportion to the inversion's value because the foundational engineering is already done. Buyers pay for customisation against their function, not for the engine to be rebuilt each time.
A custom agentic system that takes 24 months has missed the buyer's planning horizon, costs more than the inversion will return in the medium term, and burns the COO's mandate before any value materialises. A custom system that ships in 6 to 12 months lands inside the planning horizon, costs in proportion to the value, and gives the COO a return on her mandate before her next strategic review.
The difference between 24 months and 12 is the difference between an inversion that lands and an inversion that does not.
What we do not sell.
- We do not sell Teklabs.iGentic as a product.
- We do not sell the libraries standalone.
- We do not run a partner programme.
We sell custom regulated AI delivery, accelerated by the engine we already own.
Bring your gap. Tell us where the standard is going unmet. We will listen, ask, and show you work we have shipped against gaps that look like yours.