Job Title: AI Engineer
Company: Unico Connect
Location: Kurla (West), Mumbai, India (Hybrid)
Working days: Monday - Friday
Experience: 2-4 years
About the role:
Unico Connect is an AI-native technology partner that builds custom mobile, web, and AI products for clients across multiple geographies. AI is core to how we design, deliver, and scale software for our customers, and is increasingly the substance of what those customers ask us to build. We are hiring an AI Engineer who will help us build the AI capabilities, agentic solutions, and AI-enabled platforms that ship inside live customer products.
The role suits someone who thinks quickly on solutioning, can take an ambiguous problem to a working POC in days, and has the discipline to carry that POC through to a production system with real users, real traffic, and predictable economics. You will work alongside delivery pods, product managers, and customers across active engagements, with full ownership of the AI surface area of the work.
Responsibilities:
- Solutioning and POCs: Translate ambiguous customer problems into working POCs at speed. Pick the right model, framework, and architecture for the use case, and demonstrate value early before scaling investment.
- LLM application development: Build AI features and services using LLM APIs from providers such as OpenAI, Anthropic, Google, and self-hosted open-weight models (Llama, Qwen, Mistral). Choose the right model per use case based on cost, latency, capability, and context window trade-offs.
- Agentic system design: Design and implement agentic workflows using frameworks such as LangGraph, CrewAI, AutoGen, LlamaIndex Agents, or custom orchestration. Cover tool use, planning, memory, and multi-step reasoning patterns appropriate to the problem.
- API and service development: Build production AI services and APIs using Python and FastAPI. Handle streaming responses, async processing, structured outputs, retries, and graceful degradation when models or tools fail.
- Retrieval and tool integration: Implement RAG pipelines with vector databases (Pinecone, Weaviate, Qdrant, pgvector, Chroma), embeddings, chunking strategies, hybrid search, and reranking. Integrate external tools, internal APIs, and document sources through tool-calling and MCP-style patterns.
- Cost analysis and unit economics: Model the per-request and per-user cost of every AI feature before it ships. Track token usage, prompt caching opportunities, batching, and model-routing strategies. Drive measurable improvements in unit economics for AI features in production.
- Production hardening: Add observability and tracing (LangSmith, Langfuse, OpenTelemetry), guardrails, content safety checks, prompt injection defences, and fallback behaviour. Ship features that survive real customer traffic and scale.
- Prompt engineering and evaluation: Design, test, and iterate prompts with measured outcomes. Build evaluation harnesses for accuracy, hallucination, latency, and cost. Run benchmarks across models and prompt variants before locking in a design.
- Model exploration and experimentation: Track new model releases, capability shifts, and pricing changes. Run hands-on tests to compare model behaviour on representative client tasks rather than relying on public leaderboards alone.
- Continuous learning and sharing: Read current research, vendor changelogs, and applied AI engineering write-ups. Share findings through internal demos and short writeups so the team improves together.
Requirements:
- AI solutions delivered to production at scale (mandatory). Must have personally built and shipped at least one AI solution that runs in production for real users, with sustained traffic and operational ownership. POCs, internal demos, and one-off scripts do not qualify. Be ready to discuss scale (users, requests, data volume), the architecture decisions you made, and what broke and how you fixed it.
- 2 to 4 years of professional software or AI engineering experience, with at least one production AI feature you owned end-to-end.
- Strong Python proficiency and API development with FastAPI. Comfort with type hints, async, packaging, testing, streaming responses, and authentication. Production-grade Python, not notebook-only code.
- Hands-on depth across the LLM and agent stack. Working experience with at least two of OpenAI, Anthropic Claude, Google Gemini, or self-hosted open-weight models (vLLM, Ollama, Together, Replicate). Working familiarity with at least one agent framework (LangGraph, CrewAI, AutoGen, LlamaIndex Agents) or hand-rolled equivalent. Working knowledge of RAG, embeddings, and vector databases (Pinecone, Weaviate, Qdrant, pgvector, Chroma).
- Solutioning speed and POC velocity. Demonstrated ability to move from a fuzzy problem statement to a working prototype in days, not weeks. Strong instinct for what to build first, what to defer, and what to throw away.
- Cost discipline for production AI. Ability to calculate, monitor, and optimise the cost of LLM APIs, tokens, embeddings, vector store usage, and infrastructure. Treats unit economics as a first-class engineering concern, not an afterthought.
- Comfortable in a startup-style environment. Self-directed, comfortable with ambiguity, takes ownership without being asked, ships under shifting priorities, and learns by doing rather than waiting for process.
- Continuous learning and current with the field. Tracks model releases, capability shifts, and applied AI engineering practice. Has opinions, backed by hands-on tests and benchmarks, on which models suit which workloads.
- Strong written and spoken English communication. Able to explain trade-offs to non-AI engineers, designers, product managers, and clients in plain language.
How to Apply:
Candidates are encouraged to apply directly through this LinkedIn job posting. Alternatively, resumes may be sent to [Confidential Information] with the subject line Application – AI Engineer. Shortlisted candidates will be contacted regarding next steps.
