Senior AI Engineer (Full-Stack) — Detailed Job Description Position: Senior AI Engineer (Full-Stack) Location: Remote 10 hours per day Experience Required: 5+ years Python engineering, 3+ years multi-agent systems Type: Contract ROLE SUMMARY We are seeking a highly experienced Senior AI Engineer to lead the development of production-grade multi-agent AI systems, backend services, LLM orchestration, and full-stack AI-driven product experiences. The ideal candidate possesses deep technical expertise across Python backends, multi-agent workflows, LLM integrations, RAG pipelines, multimodal processing, and frontend engineering. KEY RESPONSIBILITIES ● Design and implement scalable multi-agent architectures: supervisor patterns, orchestrators, shared memory/state, workflow dependencies, checkpointing, retries, and debuggability. ● Build agent-driven coding workflows with hooks, background tasks, and toolchains integrating AI coding tools. ● Develop high-performance Python backend services using FastAPI, async concurrency, typed schemas, and secure API gateways. ● Build distributed task processing pipelines with Celery + Redis (or equivalents) for long-running AI workloads. ● Integrate multiple LLM providers with routing, fallback logic, streaming, cost optimization, tool/function calling, and JSON-structured output handling. ● Build evaluation pipelines for LLM-as-judge, human-in-loop reviews, automated prompt regression tests, and iterative prompt optimization workflows. ● Develop agentic search and crawling workflows using Playwright/Selenium/Firecrawl with LLM-ready content extraction and error handling. ● Implement production-grade RAG pipelines: vector DBs, hybrid search, chunking, metadata/RBAC tagging, embedding optimization, and retrieval policy design. ● Build real-time streaming infrastructure using SSE/WebSockets, Redis caching layers, pre-computation and rate-limiting strategies. ● Work on PostgreSQL schema design, DynamoDB key-value workflows, and ClickHouse analytics setups for event and BI use cases. ● Implement multimodal AI features: Whisper STT, TTS, OCR, vision models, document parsing, and image generation workflows. ● Support full-stack development in React/Next.js, TypeScript, Tailwind, real-time chat interfaces, and browser extension workflows. ● Establish DevOps best practices using Docker, CI/CD pipelines, monitoring dashboards, and containerized deployments. MUST-HAVE SKILLS ● 5+ years Python (async, concurrency, FastAPI, Pydantic). ● 3+ years multi-agent workflow development (LangGraph or equivalent). ● Strong expertise in Celery + Redis or equivalent distributed compute frameworks. ● Multi-LLM integration experience across OpenAI, Anthropic, Google, xAI. ● Proven RAG, vector search, embedding pipelines, and retrieval implementation experience. ● Strong background in web crawling/automation using Playwright/Selenium. ● Experience in building streaming endpoints and caching layers. ● Solid data engineering experience with Postgres, DynamoDB, and ClickHouse. ● Frontend engineering in React/Next.js + TypeScript. NICE-TO-HAVE SKILLS ● Experience with Whisper, ElevenLabs, multimodal vision systems. ● Experience with GEPA-style prompt optimization loops. ● Experience with browser extension development for AI product workflows. ● Kubernetes familiarity and advanced MLOps methodologies. SUCCESS METRICS ● High reliability multi-agent workflows with automated recoverability. ● Efficient and deterministic RAG retrieval systems. ● Low-latency streaming architecture supporting real-time AI UI. ● Successful multi-LLM orchestration with cost reduction and fallback stability. ● Production-ready evaluation and regression testing frameworks. SUBMISSION REQUIREMENTS ● CVs with highlighted relevant project work. ● Examples or case studies of multi-agent systems built. ● Details on LLM integration experience.