AI-assisted Avaloq engineering
Using GitHub Copilot and a retrieval-based knowledge base over Avaloq documentation to speed up development and functional analysis.
- 1,000+
- pages of Avaloq docs indexed
Context
Avaloq knowledge is spread across long configuration guides and language references. Answering “how is this parameterised?” or “what does this script construct do?” often means a lot of manual searching.
What I did
- Built a retrieval-based (RAG) knowledge base over 1,000+ pages of Avaloq configuration and language-reference documentation, so functional and scripting questions get answers grounded in the docs.
- Use GitHub Copilot and LLM-based tooling in day-to-day Avaloq work to speed up development and go deeper in functional analysis.
Why it matters
Answers come from the documentation rather than memory, which shortens analysis time and makes it easier to check that a proposed configuration matches documented behaviour.