We are looking for a sharp, fast-learning AI Engineer to build autonomous agent pipelines at a generative-AI startup in Gurgaon. You will work across multiple products, integrate APIs, and ship production systems that automate business processes.

About the Company

TrueFan AI is a generative-AI platform based in Gurgaon, India, that builds AI-powered celebrity video ads and hyper-personalized marketing solutions for Indian brands. The company sits at the intersection of GenAI, voice, video, and automation, shipping production systems that talk, sell, and create at scale. They focus on building autonomous, agentic products across different parts of the business – from sales and lead generation to content and localization.

Role Overview

We are looking for an AI Engineer who can move fast – integrating tools and APIs, wiring up agent pipelines, and shipping working systems quickly. This is a junior-to-mid role, and we care more about sharp thinking, ownership, and a fast learning curve than years of experience. You will work closely with the founding team on both the technical build and the underlying process/strategy for whichever product you are on. You will move across products as needed, and the common thread is always: take a business process that currently needs a human, and turn it into a reliable agent pipeline.

Key Responsibilities

• Work across multiple agent products, each automating a different business process end-to-end.
• Help define the business logic behind each agent – what 'good' looks like for a given task – and translate that into concrete, testable rules the agent can follow.
• Build and maintain data connectors and integrations – CRMs, external registries/APIs, publishing or delivery targets, and whatever each product needs to talk to.
• Own workflow and task orchestration – background jobs, queues, and webhook callbacks that move something from raw input to finished output without manual intervention.
• Iterate based on real outcomes – track versions of configs/prompts, compare results across versions, and help the team understand why a pipeline's output quality moves.
• Debug pipeline issues end-to-end – a stuck job in the queue, a malformed payload hitting a downstream API, a job that silently failed halfway through a batch.
• Work with practical LLM tooling – using an LLM to turn plain-language criteria into structured queries or outputs, and using agent tools for research/drafting/QA.

Tech Stack

Our stack is representative and evolving, and includes:

• Languages & Frameworks: Python, FastAPI, SQLAlchemy, Next.js/React (frontend team)
• Task Queues: Celery + RabbitMQ
• Agent Frameworks: LangGraph
• Database: MySQL
• Containerization: Docker
• LLM Usage: via API for structured output generation, content, and QA tasks

This is a builder's role at a fast-moving AI company – you will ship things that go live and get used the same week you build them.

What We're Looking For

• 0–3 years of experience, with solid Python fundamentals (can read and write real backend code, not just notebooks).
• Working knowledge of LLMs and how to build with them – prompting, structured output, function/tool calling, and agent patterns (multi-step reasoning, tool use, hand-offs).
• Comfortable working with APIs, databases, and background job/queue systems (even if you haven't used our exact stack before).
• Good at quickly integrating third-party tools and APIs, and getting a working pipeline shipped fast rather than over-engineering it.
• Know how to evaluate agent/LLM output – setting up evals, comparing prompt/config versions, and catching regressions before they hit production.
• Understanding of RAG – how retrieval pipelines are built, chunking/embedding basics, and when RAG is (and isn't) the right tool.
• Understand what it takes to move an agent from 'working in a notebook/demo' to running reliably in production – error handling, retries, monitoring, and graceful failure.
• Comfortable switching context across different products/pipelines, and comfortable with ambiguity – a lot of what we're building doesn't have a playbook yet, so you'll help write it.
• Genuinely willing to learn and put in the work – this is a problem-solver's role more than a 'follow the spec' role.
• Strong communication – you'll regularly need to translate between 'what the business wants' and 'what the system currently does.'

Nice to Have

• Exposure to FastAPI, SQLAlchemy, or any Python web backend.
• Familiarity with task queues and workers (Celery, RabbitMQ, or similar).
• Any prior exposure to LangChain/LangGraph or agent frameworks – helpful but not required.
• Interest in business process automation, content workflows, or CRM/sales strategy.
• Frontend exposure (React/Next.js) is a plus but not required.

Benefits

• Health insurance
• Opportunity to work closely with the founding team
• Fast-paced, builder-first environment with real impact from day one