Overview
The Agentic AI Platform is Quanta GenAI's orchestration framework. Configure models, skills, agents, and workflow graphs in Agentic Studio, then run them from the workspace to develop products across PDLC, SDLC, or custom domains.
What makes it Agentic
- Studio-first — define packs in the UI (no code deploy required for custom stages)
- Multi-agent graphs — each node binds an agent with its own model, skills, and tools
- ReAct loop — agents plan, call tools, verify results, and iterate
- Persistent pack state — pack, node, and completed stages are saved per project
- JSON import/export — share workflow packs between accounts
Flagship workflow: PDLC
Seeded system pack with six stages:
| Stage | Agent role | Primary artifact |
|---|---|---|
| Discover | Product discovery | docs/DISCOVER.md |
| Design | Solution design | docs/DESIGN.md |
| Develop | Build | Application source files |
| Test | QA & fixes | docs/TEST_REPORT.md |
| Deploy | Release | Preview URL |
| Iterate | Feedback & backlog | docs/ITERATE.md |
Clone the system pack in Studio to customize stages, agents, or add SDLC gates.
Try it
- Studio —
/studio— models, skills, agents, graph editor (login required) - Product page —
/products/agentic-ai— public PDLC demo - Workspace — pack picker + stage strip in AI Chat
- API —
/studio/api/*,/agent/stream,/agent/demo
Beyond PDLC
Cyber Security and Product Research ship as stub packs. Build ops, compliance, content, or full SDLC packs in Studio on the same GraphRunner.
Next steps
- Getting Started — Studio + workspace
- Agent Workflows — graphs and multi-agent handoffs
- Architecture — packs, GraphRunner, ReAct