Job Title: AI / Generative AI Engineer
Location: Bangalore (On-site / Hybrid)
Experience: 1 Year
Compensation: ₹15,00,000 – ₹20,00,000 LPA
Employment Type: Full-Time
About the Role: We are a cutting-edge technology startup building next-generation AI products. We are looking for a highly exceptional, early-career Generative AI Engineer to join our core AI team. In this high-impact role, you will move beyond basic API wrappers to design, build, and optimize custom LLM workflows, RAG pipelines, and intelligent AI agents.
Key Responsibilities
- Design and implement complex Retrieval-Augmented Generation (RAG) pipelines for enterprise data.
- Integrate and evaluate foundational models (OpenAI, Anthropic, Gemini) and open-source LLMs (Llama 3, Mistral) for specific product use cases.
- Build and orchestrate AI agents using frameworks like LangChain, LlamaIndex, or AutoGen.
- Develop robust backend Python APIs (FastAPI) to serve AI models to frontend applications.
- Set up and query Vector Databases (Pinecone, Qdrant, Chroma, or Weaviate) for semantic search.
- Optimize system latency, prompt architectures, and token economics for production workflows.
Must-Have Qualifications
- Minimum 1 year of hands-on, professional software engineering experience (or highly exceptional academic/internship experience in GenAI).
- Expert-level proficiency in Python.
- Demonstrable production or deep open-source project experience with LLMs, RAG, and Vector Databases.
- Experience with LangChain, LlamaIndex, or similar orchestration frameworks.
- Strong understanding of data structures, algorithms, and software engineering best practices.
Good-to-Have Skills
- Degree from a Tier-1 engineering institution (IIT, NIT, IIIT, BITS) is highly preferred.
- Experience fine-tuning open-weight models (LoRA, QLoRA).
- Familiarity with AWS/GCP and Docker deployments.
Skills: chroma,llm,rag,databases,gemini,fastapi,autogen,agents,pinecode,models,weaviate,ai,lllmaindex,agentic ai,langchain,python,vector,anthropic,openai,artificial intelligence,pipelines,qdrant,generative ai