GenAI / AI Agent Developer

Build production-ready Generative AI applications using Large Language Models (LLMs) and AI agents. Implement RAG systems, prompt engineering, and multi-agent workflows for intelligent automation. Key Responsibilities • Design and implement LLM-powered applications and AI agents • Build RAG (Retrieval Augmented Generation) systems with vector databases • Develop advanced prompt engineering strategies and templates • Create multi-agent systems with tool integration and orchestration • Implement document processing pipelines and knowledge base ingestion • Optimize LLM inference for cost, latency, and quality • Integrate LLMs with business workflows and APIs • Evaluate LLM outputs and implement guardrails and safety measures Required Skills LLM & Generative AI: • Deep understanding of Large Language Models (GPT, Claude, Llama) • Prompt engineering techniques (zero-shot, few-shot, chain-of-thought) • RAG architecture and implementation patterns • Context management and token optimization • Fine-tuning and parameter-efficient methods (LoRA, QLoRA) • Understanding of transformer architecture and attention mechanisms AI Agents & Orchestration: • Agent frameworks and autonomous systems • Tool calling and function integration • Multi-agent communication and coordination • Planning, reasoning, and reflection patterns • Memory management for conversational AI Vector Search & Embeddings: • Embedding models and semantic search • Vector database operations and optimization • Similarity search and retrieval strategies • Chunking strategies and document preprocessing Required Tech Stack LLM Frameworks & APIs: • LLM Providers: OpenAI API (GPT-4, GPT-3.5), Anthropic (Claude), OpenRouter • Frameworks: LangChain, LlamaIndex, LiteLLM, Haystack • Agent Frameworks: AutoGPT, LangGraph, CrewAI etc • Open Source LLMs: Llama 3, Mistral, Mixtral (via HuggingFace) etc Vector Databases & Search: • Vector DBs: Pinecone, Weaviate, Chroma, Qdrant, Milvus (any of one) • Embeddings: OpenAI Embeddings, Sentence Transformers, Cohere • Search: Elasticsearch, OpenSearch Development Tools: • Languages: Python (expert) • Web Frameworks: FastAPI, Flask, Streamlit • Document Processing: LangChain Document Loaders, Unstructured, PyPDF2 • NLP Libraries: spaCy, NLTK, Hugging Face Transformers MLOps & Deployment: • Model Serving: vLLM, Ray Serve, TGI (Text Generation Inference) • Monitoring: LangSmith, Weights & Biases • Containerization: Docker, Kubernetes • Version Control: Git Cloud & Infrastructure: • Cloud Providers: AWS (Bedrock, Lambda), Azure (OpenAI Service), GCP • APIs: REST, WebSocket, GraphQL • Caching: Redis Preferred Qualifications • Bachelor's/Master's in Computer Science, AI, NLP, or related field • Experience fine-tuning LLMs (LoRA, full fine-tuning) • Knowledge of LLM evaluation frameworks (ROUGE, BLEU, BERTScore) • Contributions to LLM/GenAI open-source projects • Experience with multi-modal models (vision, audio) What Success Looks Like • Production GenAI applications handling real user traffic • High-quality LLM outputs with low hallucination rates • Cost-optimized inference with acceptable latency • RAG systems providing accurate, relevant context • Robust agent systems completing complex multi-step tasks • Well-structured prompts and reusable templates