04 / GenAI for decision support
PreCast
From municipal indicators to AI-assisted analysis.
A municipal-management MVP combining dashboards, contextual chat, and GenAI-generated analysis and recommendations.
- Context
- Municipal management · Product in development
- Status
- MVP / prototype
- Technology & practices
- Azure OpenAIVercel AI SDKNext.jsPostgreSQL
- Access
- Technical contribution overview. No public demo linked.
01
The context
Municipal teams need to interpret indicators across service areas and regions. PreCast explores how to bring that work into everyday workflows through dashboards, natural-language questions, and analysis generated from supplied data.
02
My contribution
- Developed dashboards and journeys by service area and district, with AI features integrated into the interface.
- Integrated Azure OpenAI through the Vercel AI SDK for chat, streaming responses, and recommendation generation.
- Structured ObservaSampa indicator ingestion and PostgreSQL storage for data and generated analysis.
03
What was built
An MVP with interfaces and APIs to explore indicators, use a contextual assistant, and save AI-generated analysis. Recommendation generation takes supplied data as input; chat uses the screen context. Time-series forecasting is a separate statistical feature.
A similar challenge in your team?
Let’s discuss your needs and my experience with this kind of project.