Unico Connect
Website:
unicoconnect.com
Job details:
Junior AI EngineerLLM Applications, RAG & APIs
Mumbai (On-site) | Full-time | 1-2 years
About the roleUnico Connect is an AI-first technology partner that builds custom mobile, web, and AI products for clients across multiple geographies. We are hiring a Junior AI Engineer who will work alongside our AI engineers to build LLM-powered features, prototype solutions for customer problems, and contribute to production AI services.
The mandatory requirement for this role is hands-on Python and LLM API experience, demonstrated through at least one substantive project that goes beyond a tutorial or coursework. Portfolio projects, internships, hackathon work, or contributions to open-source AI projects all qualify if they show real depth. The role is hands-on. Expect to build prototypes, write production-grade Python services, work on RAG pipelines, run evals, and learn rapidly from senior AI engineers on the team. A typical week includes pairing with a senior engineer on an active feature, a solo build of a prototype for a new use case, and a research spike on a new model or technique.
Responsibilities- Prototype and POC development: Build prototypes for customer use cases using LLM APIs and vector databases. Move quickly, demonstrate value, and iterate.
- AI service development: Contribute to production AI services and APIs using Python and FastAPI. Cover request handling, validation, structured outputs, and error handling.
- RAG pipeline implementation: Build retrieval pipelines with embeddings, vector databases (Pinecone, Weaviate, Qdrant, pgvector, Chroma), chunking strategies, and reranking.
- Prompt engineering and evaluation: Design and iterate prompts. Build evaluation cases and measure model behaviour across changes.
- Cost and performance awareness: Track token usage, latency, and cost on the features you work on. Flag inefficiencies to senior engineers.
- Observability: Instrument AI workflows with LangSmith, Langfuse, or equivalent. Use traces to debug and improve outputs.
- AI-assisted development: Use Claude, Cursor, and similar tools day to day. Build the discipline to review and validate AI-generated code before shipping.
Requirements- Hands-on AI project experience (mandatory). Must have built at least one substantive AI project beyond tutorials or coursework. Internship work, hackathon projects, side projects, and open-source contributions qualify if they demonstrate real depth and ownership. Be ready to walk through the project, the choices you made, and what you would do differently.
- 1 to 2 years of professional software development experience. Exceptional candidates with strong project portfolios but slightly less professional time are welcome to apply.
- Strong Python proficiency. Type hints, async, packaging, testing. Production-quality code, not just notebooks.
- Working knowledge of LLM APIs. Hands-on with at least one of OpenAI, Anthropic Claude, or Google Gemini. Comfortable with prompts, structured outputs, and streaming responses.
- RAG fundamentals. Practical experience with embeddings, vector databases, and retrieval pipelines.
- REST APIs and FastAPI basics. Built at least one API service in Python.
- SQL and Git fundamentals. Comfortable with relational databases and version control workflows.
- Strong written and spoken English. Able to explain technical work clearly in writing and in conversation.
- Curiosity and ownership. Reads, experiments, asks questions, and takes initiative.
Nice to have: at least one agent framework (LangGraph, CrewAI, LlamaIndex Agents); fine-tuning exposure; AWS familiarity; open-source AI contributions; technical write-ups or blog posts on AI topics.
Click on Apply to know more.