Full Stack AI Engineer

Appinessworld; Pune District; On-site; INR 2,500,000 - 4,500,000; Full time

Job summary

Appinessworld is seeking a Full Stack AI Engineer in Pune to own end-to-end feature delivery, from backend agent orchestration to intuitive frontend interfaces for configuring, monitoring, and debugging agent workflows.

You will work with architects and product owners to translate orchestration concepts into a production-grade system, shipping agents, APIs, and UI with strong emphasis on reliability, testing, and scalability.

Qualifications

• Must have 5+ years of full-stack experience with Python (FastAPI) and React.
• Hands-on with at least one agent framework: AG2 AutoGen, LangGraph, LangChain Agents, or CrewAI.
• Experience in LLM application development: prompts, tool calling, streaming responses, and context management.

Responsibilities

• Backend & Agent Development: design multi-agent workflows and build FastAPI services.
• Frontend Development: build React interfaces for configuring agents and visualising conversations.
• System & Production: containerise with Docker, contribute to CI/CD, implement logging and tracing.

Skills

Python

FastAPI

React

AG2 AutoGen

LangGraph

LangChain Agents

LLM Apps

RAG Embeddings

PostgreSQL

Async Patterns

Docker

CI/CD

WebSockets

Tools

pgvector

Qdrant

Weaviate

FAISS

Ollama

vLLM

LangSmith

Job description

# Full Stack AI Engineer* Careers* Full Stack AI EngineerLocation: PuneRequired Experience: 4-7 Years## Job DescriptionWe are looking for a Full Stack AI Engineer who can own end-to-end feature delivery – from designing agent communication patterns and orchestration logic on the backend, to building intuitive interfaces for configuring, monitoring, and debugging agent workflows. You will work directly with the architect and product owner to translate orchestration concepts (agent-to-agent messaging, tool use, handoffs, memory) into a production-grade system.This is a hands-on builder role. You will not be writing strategy decks – you will be shipping agents, APIs, and UI.## Roles & Responsibilities:**Key Responsibilities****Backend & Agent Development*** Design and implement multi-agent workflows using AG2 (AutoGen) and/or LangGraph – including agent roles, conversation patterns, handoffs, and termination conditions.* Build and maintain FastAPI services that expose orchestration capabilities (agent registration, task submission, run status, streaming responses) to the frontend and external consumers.* Implement tool/function calling, structured outputs, and memory layers (short-term, long-term, vector-based) for agents.* Integrate LLM providers (OpenAI, Anthropic, AWS Bedrock, or local models via Ollama) with proper retry, timeout, and cost controls.* Write clean, testable Python with type hints, pydantic models, and clear separation between orchestration logic and infrastructure.**Frontend Development*** Build React-based interfaces for configuring agents, visualising agent conversations, inspecting tool calls, and debugging runs.* Implement streaming UI (SSE/WebSockets) to render token-by-token agent responses and intermediate steps.* Collaborate on UX for workflow builders, run history, and observability dashboards.**System & Production Concerns*** Containerise services with Docker and contribute to CI/CD pipelines.* Add logging, tracing, and basic evaluation hooks for agent runs (latency, token usage, success rates).* Participate in code reviews, design discussions, and incremental hardening of the platform.## Pre Requisites:**Must-Have Skills*** Full stack experience (5+ years): strong Python (FastAPI or similar) on the backend AND React on the frontend. Comfortable owning a feature across both layers.* Hands-on with at least one agent framework: AG2 (AutoGen), LangGraph, LangChain Agents, or CrewAI. Must have actually built and run multi-agent flows, not just read about them.* LLM application development: prompt design, tool/function calling, structured outputs, handling streaming responses, and basic context management.* RAG fundamentals: chunking strategies, embeddings, vector stores (pgvector / Qdrant / Weaviate / FAISS), and retrieval evaluation. Bonus for hybrid (sparse + dense) retrieval.* Database skills: PostgreSQL – schema design, indexing, and writing reasonable queries.* API & async patterns: REST design, async Python (asyncio), background jobs, and streaming endpoints.* Version control & collaboration: Git, PR-based workflows, writing clear commit messages and design notes. **Good-to-Have Skills*** Experience with GCS for deploying AI workloads.* Familiarity with LLMOps – tracing (LangSmith / Langfuse / OpenTelemetry), evaluation frameworks, and prompt versioning.* Exposure to MCP (Model Context Protocol) or A2A (agent-to-agent) communication standards.* Experience running local LLMs (Ollama, vLLM) for prototyping or cost-sensitive workloads.* Knowledge of WebSockets / SSE for real-time UIs.* Basic understanding of fine-tuning concepts (LoRA / QLoRA) – we are not expecting trainers, just awareness.## Skills:PythonReactFast APIPostgreSQLGen AI & Multi Agent