//fiLAB.ai
U imagine IT, we create IT
fiLAB.ai is an IT and R&D consultancy run by working engineers — a decade of shipping systems for auditing, finance, logistics, telecoms and research.
In the open
Client work stays confidential. What's on the workbench right now is public — five projects, from proof automation to trading infrastructure.
//For IT leadership
Architecture that serves the business, not the diagram
Advice and leadership for organisations whose IT estate has outgrown its original design.
Architect as a service
We sit inside your team as architect and tech lead: we set the target state, make the technical calls, and argue with vendors and partners on your behalf.
Engineering & product leadership
We lead engineering teams day to day — and take the product ownership with them, from process analysis through every delivery.
Cloud & SRE
We move systems onto AWS, Azure, GCP or Cloudflare, then run them with reliability discipline — from monitoring to a staffed on-call rota.
Platform architecture & APIs
We design and operate Kubernetes platforms — platform-as-a-service included — and shape domains into service mesh, REST, GraphQL, pub/sub and socket APIs.
//For teams that need capacity
Software engineering & data science contracting
We plug in as engineers — from data science to ML/AI-enabled products.
Full-stack product engineering
We build web, mobile and back ends — TypeScript, React, Vue, Astro, Kotlin, Flutter, Swift, Qt and OpenGL on the client; Spring, ZIO, Akka, Django and two decades of Python on the server.
Data & ML/AI platforms
We assemble streaming and batch pipelines on ClickHouse, Kafka, Flink, Spark, Airflow, Iceberg and Postgres — and keep the data governable with models, documentation and visualisation people actually use.
High-performance & GPU code
We write C, C++, Rust, Fortran and assembly for low-latency and parallel workloads — LLVM/MLIR, CUDA, OpenCL.
Multimodal AI
We build retrieval-augmented generation, computer vision and speech-to-text — plus symbolic AI over knowledge bases.
Compilers, parsers & graph analysis
We extend compilers such as LLVM, and write parsers together with complex graph and semantic analysis.
Salesforce & GIS integrations
We integrate where the business already lives: ESRI, MicroStation and the open-source GIS stack, and Salesforce with its UI framework, Apex and Flow.
//For open research problems
Research & development
Product R&D
We take data and software products from prototype — or an unclear problem space — to a validated build in production.
Language design & implementation
We design programming languages and extend compilers such as LLVM.
Algorithms & symbolic computing
We design algorithms and data structures, drive MATLAB, Maple and Mathematica where standard tooling gives up, and build rule engines, reasoners and optimisation solvers.
How the work starts
The same three steps every time — no procurement maze:
Discovery call
Bring the biggest open question. Half an hour is usually enough to see whether it fits.
Scoped pilot
One defined slice of work, clear exit criteria, a usable result — weeks, not quarters.
Retained or delivered
Long-running advisory, or end-to-end delivery — documented, reviewed and handed over.
Let's talk about the next build
Bring scope and constraints — you get a concrete next step in return.