Unico Connect
Website:
unicoconnect.com
Job details:
Full Stack Engineer
Python, FastAPI, React & AI-Native Architectures
Mumbai (On-site) | Full-time | 2-4 years
About the role
Unico Connect is an AI-first technology partner that builds custom mobile, web, and AI products for clients across multiple geographies. We are hiring a Full Stack Engineer who will work on customer engagements, building AI-native architectures, FastAPI services, and React interfaces that ship in front of real users.
The mandatory requirement for this role is hands-on production experience building Python services with FastAPI on the backend and React on the frontend, with at least one shipped product or platform that runs for real users. The role is in-office at our Mumbai office and is customer-facing, with direct interaction during discovery, design reviews, demos, and project conversations. The work is hands-on across the stack with a backend lead: designing and building FastAPI services, integrating LLMs and AI workflows into the product, owning data models, and shipping React interfaces where the product calls for it. A typical week includes building a new FastAPI endpoint with LLM integration, refining a React flow, running a working session with a customer, and reviewing your own production deployment.
Responsibilities
Backend development with Python and FastAPI: Design and build production APIs and services using Python and FastAPI. Cover async patterns, streaming responses, structured outputs, authentication, retries, and graceful degradation.
AI-native architecture and integration: Design product architectures where LLMs, retrieval pipelines, agents, and AI workflows are first-class components. Integrate LLM APIs from providers such as OpenAI, Anthropic, and Google, and build the orchestration, evaluation, and observability around them.
Frontend development with React: Build React interfaces (with Next.js where relevant) for the products you ship. Own the end-to-end flow from API to UI for the features in your scope.
Data modelling and database design: Own schema design, indexing, query performance, and migrations on PostgreSQL. Make and defend data model decisions based on access patterns and product needs.
Customer engagement: Work directly with customers during discovery, design reviews, demos, and weekly working sessions. Translate fuzzy product asks into concrete technical proposals and ship the work.
POC to production discipline: Move from a rough product idea to a working POC in days, then carry the POC through to a production system with real users, real traffic, and predictable cost.
Cost and unit economics for AI features: Track token usage, latency, and per-request cost for the AI features you ship. Apply caching, batching, and model routing where the workload justifies it.
AI-assisted development: Use Claude, Cursor, and similar tools day to day. Develop strong instincts for prompts, patterns, review, and verification of AI-generated code.
Code quality and reviews: Write tests, run code reviews, follow branching and release conventions, and raise the quality bar on the codebases you contribute to.
Continuous learning: Track new model releases, framework updates, and applied AI engineering practice. Share findings through internal demos and short write-ups.
Requirements
Hands-on production experience with Python (FastAPI) and React (mandatory): Must have personally built and shipped at least one production application using FastAPI on the backend and React on the frontend, with real users and sustained traffic. POCs, coursework, and internal tools do not qualify.
2 to 4 years of professional full stack engineering experience, with at least one production system you owned end-to-end. Candidates with slightly less experience but demonstrably strong ownership are welcome to apply.
Strong Python proficiency: Comfort with type hints, async, packaging, testing, dependency management, and production patterns. FastAPI in production, not framework demos.
Working knowledge of LLM and AI engineering patterns: Hands-on integration with at least one of OpenAI, Anthropic Claude, or Google Gemini in a production or near-production context. Familiarity with prompt design, RAG, embeddings, and vector databases (Pinecone, Weaviate, Qdrant, pgvector, or Chroma).
Production-grade React with TypeScript: Comfort building real interfaces, managing state, and consuming APIs cleanly. Working familiarity with Next.js is preferred.
PostgreSQL skills: Schema design, indexing, query performance, and migrations on at least one production system.
Cloud and deployment fundamentals on AWS, GCP, or Azure: Hands-on with deploying and operating services, configuring CI/CD pipelines, and basic observability.
Bachelor's degree in Computer Science, Information Technology, or a related engineering discipline: Exceptional candidates with demonstrable production experience and strong portfolios may be considered without a formal degree.
Strong written and spoken English communication: This role is customer-facing. Comfort presenting technical trade-offs, running working sessions, and writing clear status and design documents is essential.
Nice to have: experience with agent frameworks (LangGraph, CrewAI, AutoGen, LlamaIndex Agents); evaluation tooling (LangSmith, Langfuse, Promptfoo, Ragas); Docker and Kubernetes in production; published technical write-ups or open-source contributions; prior agency, consulting, or product-engineering experience.
Click on Apply to know more.