Full Stack AI Engineer

Full-Stack AI Engineer (Intern or Full-Time)

Stack: Java/Spring (backend), React (web), Flutter (mobile), MCP, Python/Node, Docker

Location/Mode: Gurgaon • Onsite/Hybrid

Role

Ship production AI agents (chat + voice) tightly integrated with our Java APIs via MCP. Own the pipeline end-to-end: prompts → tools → RAG → evals → deploy → observe.

Responsibilities

• Expose Java/Spring endpoints as MCP tools/servers; wire agents to use them safely.

• Build LLM agents (workflow/chat/voice IVR) with streaming UX in React/Flutter.

• Voice stack: ASR (Whisper/Cloud) + TTS + real-time audio.

• Data layer: embeddings, vector DB (pgvector/Pinecone/Weaviate), caching, rate-limits.

• Evals & guardrails: latency/cost metrics, prompt/versioning, safety/PII handling, logs/traces.

Must-Haves

• Evidence of shipped agents (links/GitHub): tool/function-calling, retrieval, or voice bot.

• Practical RAG & embeddings; REST/WebSockets; Docker; basic CI/CD.

• Comfortable reading Java/Spring; one runtime for AI services (Python or Node).

• SQL (Postgres) and queues/batch basics.

Nice-to-Have

• Direct MCP experience; local models (Ollama); OpenTelemetry; AB tests/feature flags.

• Flutter/React chat or voice component integration.

What to Submit

• Resume + GitHub + 1–2 live demos.

• 5-line summary of your most complex agent: architecture, evals, p95 latency, cost/session, and which Java endpoints/tools (or MCP design) it used.