We are thrilled to announce a major breakthrough in our development: the core base environment for the AICICD Autonomous Engineering Platform (V3) is officially up and running! This stable foundation is now fully set up to implement and test our highly anticipated Proof of Concept (POC).The Architecture: Orchestration meets RuntimeOur platform achieves true autonomy by separation of concerns, splitting the workload into two powerful layers that are now functional in this base environment:
- AICICD (The Orchestrator): Manages the overall project lifecycle, from project bootstrapping and workflow execution to tracking Baselines, Deltas, and Snapshots.
- KAgent (The Runtime): Provides the intelligent agent execution engine. It natively supports advanced frameworks like LangGraph and CrewAI, manages agent memory, and routes tools seamlessly via the modern Model Context Protocol (MCP).
Redefining the “Engineering Truth”One of the most exciting innovations included in this environment is our data strategy. Unlike traditional systems, Git is not our source of truth. While Git safely stores our source code and documentation, the actual Engineering Truth-including requirements, system architecture, decisions, and traceability-is entirely managed and stored in a powerful graph database (ArcadeDB).What happens next?With the infrastructure ready, we are now executing our automated Project Creation Workflow. The system is ready to automatically validate templates, provision code repositories via Gitea, initialize vector knowledge spaces via Qdrant, and orchestrate specialized AI agents (like our Chief Engineer and Technical Reviewer) to work together autonomously.This architecture brings us a massive step closer to our goal of creating a fully automated, AI-driven engineering platform. Stay tuned as we run our upcoming POC demos and share the results right here!
#supported by ai