Roadmap
What we're
building next.
ElectriPy AI is implementing the full LSAS stack — nine layers of runtime infrastructure for production AI systems. Here is where we are and where we are going.
Now
Foundation Layers
Core runtime SDK — pip install electripy-ai
Reliability layer — circuit breakers, retries, fallback routing
Observability layer — OpenTelemetry tracing, structured spans
Governance layer — policy hooks, redaction, approval gates
Model abstraction — provider-agnostic LLM gateway
Next
Application Layers
Orchestration layer — routing, coordination, planning primitives
Memory layer — session state, history, context windows
Knowledge layer — retrieval, RAG, embedding interfaces
Tools layer — MCP support, skill registry, function health
Evaluation framework — automated scoring, regression suites
Later
Cloud Platform
ElectriPy Cloud — hosted runtime with managed LSAS stack
Trace explorer — production session replay and drill-down
Policy console — team-level governance controls
Deployment targets — cloud, edge, on-prem
Enterprise licensing and SLA tiers
Follow along on GitHub.
The roadmap is tracked publicly on GitHub. Issues, milestones, and discussions are open. Community contributions are welcome.
Start building now.
The foundation layers are available today. Open source, MIT licensed.