Agent Marketplace
All demos are live, governed by shared platform infrastructure — guardrails, rate limiting, observability, and eval-gated CI. Browse by capability tier or explore the full catalog.
Part of the Enterprise Capability Lifecycle — this page covers Execution.
Governance layer
Shared
Rate limiting
All routes
Eval gating
CI-enforced
Audit trail
Trace IDs
Governed Agent Platform
FlagshipOne synthetic finance-ops task end to end on real protocol endpoints: an A2A v1.0 agent, an MCP server, a policy-enforci…
RAG Pipeline
Retrieval-augmented generation with Transformers.js embeddings (all-MiniLM-L6-v2) over an in-memory vector index — runs …
LLM Router
Multi-model routing across fast and large Groq-hosted models — see live latency, cost, and quality trade-offs per reques…
AI Evaluation Showcase
EvaluationEvaluation pipeline walkthrough — deterministic rubric scoring (required-coverage and forbidden-topic checks), guardrail…
Multi-Agent System
AgenticThree specialized agents — Analyzer, Researcher, and Strategist — with real LLM calls via Groq, handing off in sequence;…
MCP Tool Demo
ProtocolMCP-style tool use with per-tool authorization — watch an LLM select and call JSON-Schema-defined tools (via Groq functi…
AI Portfolio Assistant
Streaming full-context assistant over my experience with optional retrieval-enhanced grounding and cited context cues.
Model Quantization
Live ONNX benchmark comparing INT8 vs FP32 inference — real file sizes, real latency, real quality diff.
Enterprise Control Plane
EnterpriseOrg-wide AI governance dashboard — RBAC, group spend limits with token-cost tracking, and structured observability feed …
Edge Agent + Cloud Agent Collaboration
Edge + CloudThree-tier privacy-first AI pipeline: BERT NER redacts PII in the browser via Transformers.js ONNX, the server re-checks…
Agent Auth Demo
IdentityLive auth.md protocol implementation — AI agents register anonymously, claim with email + OTP, then call MCP tools with …
STORM Research Agent
Multi-perspective AI research inspired by Stanford STORM — discover expert lenses, generate questions, research each ang…
Real-Time Spatial AI + World Modeling Engine
Perception → reconstruction → agent reasoning. Precomputed 3D mesh playback with drift correction visualization and LLM …
Constrained Generative UI
Ask about Prasad's background — the model can only emit JSON for a fixed component catalog, validated server-side before…
Signature Agent — Start Here
Governed Agent Platform
The flagship: a real A2A v1.0 agent and MCP server behind a policy-enforcing tool gateway, with human approval, poisoned-data defense, tracing, and a trajectory-evaluation release gate.
Explore Flagship Agent