Contact

EXPERTISE

Generative AI & agentic systems

Integrating AI and agents into information systems with explicit rules for data, context, identities, tools, permissions, Runtime, testing, control and observability.

Discuss your system

Capabilities

Capabilities

From business context and decisions through to production: understand, architect, build, secure, industrialize and evolve without artificially separating information systems, software, data, automation and AI.

  1. RAG

  2. Context Engineering

  3. Multilingual assistants

  4. Agents

  5. Orchestration

  6. MCP

  7. Tool calling

  8. Agent identity

  9. Runtime

  10. Model gateway/routing/fallback

  11. Evaluation

  12. Verification

  13. Security

  14. Audit

  15. Observability

  16. LLMOps

  17. Serving

  18. French/Arabic/Darija depending on the use case

Engineering approach

Designing what must remain under control.

Code, rules, Data or AI: technical choices serve the system’s needs.

Discover our approach

Engineering approach

The model is chosen according to quality, cost, latency, privacy, languages, on-premises requirements and operational constraints. Where possible, the architecture supports comparison, fallback and switching models.

OUR ENGINEERING

  1. Problem first

    Decisions, processes, users, data, constraints and responsibilities before tools.

  2. Production by design

    Security, performance, availability, scalability, CI/CD, observability, rebuilding and maintainability shape the architecture from the outset.

  3. Evidence before promise

    Tests, metrics, traces, acceptance criteria, Evaluation, Verification and regression testing provide evidence of quality.

Technologies

Technologies

  • FAISS
  • FastAPI
  • Keycloak
  • LangChain
  • Langfuse
  • LangGraph
  • MCP
  • OPA
  • OpenTelemetry
  • Phoenix
  • Qdrant
  • Sentence-Transformers
  • Transformers
  • vLLM

Related work

Related work

  1. Agentic AI platform and multilingual assistants

    A reusable application foundation for integrating assistants with the knowledge and services of an information system.

  2. Multi-agent analytical assistant

    Connecting natural language and structured data while keeping control of queries and results.

  3. Multilingual institutional assistant

    Making services and information easier to access through conversational interaction and guided journeys.

Frequently asked questions

Frequently asked questions

Which LLM do you use?

We are not tied to a single model.

Explore

LET’S TALK

Let’s discuss your challenge.

A system to build or evolve?

Discuss your system