Generative AI Engineer | Artificial Intelligence | Mumbai

About the team

We build and run the AI systems that Square Yards' sales, marketing and operations teams use every day — about 2 billion LLM tokens a day in production. This is not a research lab and it is not a slide deck — everything we ship talks to real customers, real agents, or real listings.

Today that includes:

• A real-time AI voice calling platform that runs outbound cold-calling, lead qualification and inbound listener campaigns across multiple telephony providers, with per-campaign prompt logic, live audio streaming, barge-in handling and CRM write-back.
• A conversational intelligence stack that transcribes, diarizes and scores every call — utterance-level sentiment, interruption detection, objection mining, QA scorecards, and post-call coaching advice — exposed as a public API product.
• Retrieval-grounded assistants for property discovery, internal HR workflows, and in-product support widgets embedded directly in our dashboards.
• A generative media pipeline that turns property listings, floor plans and locality data into marketing videos and imagery at scale.
• Operational tooling — health-check services, campaign managers, analytics dashboards — that keeps all of the above observable and debuggable.

We move fast, own our systems end to end, and are small enough that one engineer's work is visible across the business.

What this role is

This is a build-and-learn role. You'll start on well-scoped pieces of live systems, and grow into owning a surface of your own. You will be writing code that runs in production within your first few weeks — not shadowing, not building sandbox demos that get thrown away.

What you'll do

• Ship features inside our live AI systems — a new analysis parameter on the call scorecard, a new campaign configuration, a new endpoint on an existing service. Small, real, reviewed.
• Write and iterate on prompts against real data. Read actual call transcripts and chatbot conversations, find where the system got it wrong, and fix it. A large share of AI engineering is this, and it is genuinely skilled work.
• Build evaluation sets and test harnesses. Assemble golden datasets, run prompt and model comparisons, and report what actually changed. You'll often be the person who knows whether a change helped.
• Build internal tools and dashboards — Streamlit, Gradio or React — so the rest of the team can inspect what a pipeline is doing without reading logs.
• Handle data plumbing: pull and clean data from MongoDB and our APIs, write pollers and background jobs, wrangle audio and transcript files into shape.
• Debug production issues — trace a bad call outcome back through the transcript, the prompt, the model response and the CRM write, and figure out which of those broke.
• Learn the stack in depth — real-time audio, retrieval, structured extraction, self-hosted inference — with the expectation that you'll be leading pieces of it within a year.

What we're looking for

Required

• Solid Python. You're comfortable with functions, classes, virtual environments, reading someone else's code, and using a debugger instead of guessing.
• You have actually built something with an LLM API — a project, an internship, a hackathon, a side project. We care that you've hit the messy parts (bad outputs, rate limits, parsing failures), not that it was at a job.
• Basic understanding of how modern LLMs work: tokens, context windows, temperature, embeddings, why the same prompt gives different answers.
• Git — branches, pull requests, reasonable commits.
• Enough SQL or database familiarity to pull the data you need without asking someone every time.
• Clear written communication. You can explain what you tried and where you got stuck without being asked three follow-up questions.
• Genuine curiosity and self-direction. Our stack changes every few months; nobody here has been handed a complete syllabus.

Strongly preferred

• Coursework, projects or internships in ML, NLP, speech, or computer vision.
• Exposure to a Python web framework — FastAPI or Flask.
• Familiarity with any of: LangChain / LangGraph, Docker, Linux command line, pandas, PyTorch.
• Some front-end ability — React, Streamlit or Gradio — enough to put a UI on your own work.
• Comfort with audio or video tooling (ffmpeg, librosa, moviepy) & with data labeling and annotation workflows.

Nice to have

• A portfolio, GitHub profile, or writeups of things you built because you wanted to see if they'd work. This counts for more with us than a certificate list.
• Open-source contributions, Kaggle work, or published projects.
• Interest in real estate, sales operations, or the business side of what the technology is for.

Tech you'll touch

Python · FastAPI · Flask · MongoDB · Docker · GCP · Google Gemini · OpenAI · Claude · Groq · BAML · LangChain / LangGraph · Pinecone · ElevenLabs · PyTorch · Transformers · OpenCV · MoviePy · React · TypeScript · Streamlit · Gradio

Why this role

Most junior AI jobs are either data labeling with a nicer title or building demos that never ship. This one isn't. You'll work on systems that make thousands of real calls and answer real customer questions, you'll see the consequences of your changes within days, and you'll be sitting next to people solving hard problems in real-time voice, retrieval and generative media — with room to grow into owning that work yourself.

Along with your resume, please include a short note on something you built with AI — what it did, what broke, and what you'd do differently. A rough side project you can talk about honestly is worth more to us than a polished one you can't!