# Full Stack AI Engineer

**Company:** [Vinmar International](https://jobs.workable.com/companies/tnLSATGhRW2vPHjMxVU7PQ.md)
**Location:** Mumbai, India
**Workplace:** on site
**Employment type:** Full-time
**Department:** Vinmar Business Services

[Apply for this job](https://jobs.workable.com/view/e70b0cf1-514c-4d16-a4d3-675547b7caea)

## Description

-   Design and build full-stack AI features, including tables, document-centric interfaces, review flows, and real-time or streaming interactions.
-   Develop agentic systems with tool calling, multi-step workflows, RAG, and structured output handling.
-   Build backend services using ASP.NET Core Web API and Python (FastAPI), and integrate them with React/Next.js frontends.
-   Implement and improve RAG pipelines, covering chunking, embedding selection, vector store integration, and retrieval quality evaluation.
-   Design AI-native UX patterns: confidence indicators, citations and source grounding, fallback states, edit/retry flows, and human review steps.
-   Write evaluation tests before shipping new AI capabilities, using golden datasets, regression gates, and CI controls.
-   Contribute to evaluation pipelines that combine deterministic metrics with LLM-as-judge approaches.
-   Build systems that degrade gracefully when model outputs are unexpected.
-   Manage context windows through token budgeting, truncation, and tool-call state persistence.
-   Prototype quickly with AI tooling, then validate production artifacts against defined quality gates before promotion.

## Requirements

**Education and experience**

-   Bachelor's degree in Computer Science, Engineering, Information Systems, Data Science, or a related field, or equivalent practical experience.
-   3+ years building production software, including full-stack applications and/or AI-enabled systems.
-   Experience contributing to user-facing AI product features, from backend through frontend.
-   Experience with agentic systems in production or pre-production (tool calling, multi-step workflows, RAG, structured outputs).
-   Exposure to evaluation frameworks such as golden datasets, regression gates, or CI quality controls.

**Full-stack product engineering**

-   Hands-on experience with .NET Core, ASP.NET Core Web API, SQL, and Microsoft technologies.
-   Frontend skills in React and/or Next.js, TypeScript, component-based UI, API integration, and state management.
-   Experience building tables, document-centric interfaces, review flows, or streaming experiences.
-   Understanding of UX patterns for AI systems (confidence, citations, fallbacks, edit/retry, human review).

**AI platform engineering**

-   Python proficiency, including FastAPI, Pydantic v2, async patterns, and pytest.
-   Hands-on experience with LangChain and/or LangGraph: stateful graphs, tool integration, checkpointing, and streaming.
-   Prompt engineering skills: structured output design, system prompt construction, and multi-turn context management.
-   Experience with RAG pipelines and familiarity with vector databases such as pgvector and/or OpenSearch.
-   Familiarity with LLM evaluation: golden dataset design, metric definition, and regression gates.
-   Awareness of context window management strategies.

**Generative AI and agentic systems**

-   Regular use of AI coding assistants (Cursor, GitHub Copilot), with good judgement about where generated code is reliable and where it needs scrutiny.
-   Familiarity with multi-agent concepts: orchestration logic, tool interfaces, and failure-handling patterns.