Lead AI Engineer

<h1 id="job-description-lead-ai-engineer">Job Description – Lead AI Engineer</h1><p>14 Oct 2026 · <span class="ldoc-mention">@Shruti</span></p><p><strong>Experience:</strong> 4–6 years · <strong>Type:</strong> Full-time · <strong>Location:</strong> Bengaluru</p><h2 id="about-the-role">About the role</h2><p>We are hiring a hands-on Lead AI Engineer to own our AI stack end to end: generative AI, agentic systems, deep learning, NLP and computer vision. You will set technical direction, lead a small team, and ship AI products that are reliable, safe and measurable in production.</p><h2 id="key-responsibilities">Key responsibilities</h2><ul><li>Architect and build GenAI and agentic AI solutions: multi-agent workflows, tool calling, memory and human-in-the-loop.</li><li>Design, optimise and evaluate RAG pipelines and vector search for accurate, grounded answers.</li><li>Fine-tune and adapt LLMs and other foundation models for domain-specific tasks.</li><li>Build and deploy deep learning, NLP and computer vision models for prediction, classification and automation.</li><li>Evaluate AI systems and optimise them for quality, cost and latency.</li><li>Implement guardrails, security and responsible-AI practices (bias, privacy, hallucination and prompt-injection controls).</li><li>Lead and mentor AI engineers, review code and designs, and partner with product and business teams on use cases and success metrics.</li></ul><h2 id="required-skills">Required skills</h2><ul><li><strong>Programming:</strong> Expert Python (NumPy, Pandas, FastAPI, async); SQL.</li><li><strong>Generative AI and LLMs:</strong> OpenAI, Anthropic Claude, Gemini, Llama and Mistral; prompt engineering, structured outputs and function calling.</li><li><strong>Agentic AI:</strong> LangChain, LangGraph, LlamaIndex, CrewAI or AutoGen; MCP (Model Context Protocol) and tool integrations.</li><li><strong>RAG and vector databases:</strong> Chunking, embeddings, hybrid search and re-ranking; Pinecone, Weaviate, Qdrant, Milvus, Chroma, FAISS or pgvector.</li><li><strong>Fine-tuning:</strong> Hugging Face Transformers, LoRA/QLoRA, PEFT and RLHF/DPO basics.</li><li><strong>Deep learning:</strong> PyTorch and TensorFlow/Keras; CNNs, RNNs and Transformers.</li><li><strong>NLP and computer vision:</strong> Text classification, NER, summarisation; OpenCV, YOLO and vision-language models.</li><li><strong>Evaluation and responsible AI:</strong> Ragas or DeepEval, guardrails, bias and safety testing.</li><li><strong>Cloud:</strong> AWS (SageMaker, Bedrock), Azure (Azure OpenAI, Azure ML) or GCP (Vertex AI).</li><li><strong>Education:</strong> Bachelor's or Master's in Computer Science, AI, Data Science or a related field.</li></ul><h2 id="nice-to-have">Nice to have</h2><ul><li>Knowledge graphs or graph RAG, speech AI (ASR/TTS), multimodal models, edge or on-device inference, and open-source contributions or publications.</li></ul>