AI Developer - OneGuru | Onsite, Noida | Open to India-based candidates only

AI Developer - OneGuru | Onsite, Noida | Open to India-based candidates only

The Details

Experience: 3+ years, with at least one AI/ML system shipped to production

Reports to: Engineering Lead

Works with: Product and a small, fast-moving engineering team

Location: Noida, India (work from office)

About the Role

We're looking for an AI Developer to build the intelligence at the core of OneGuru—our AI-native talent and skills intelligence platform. This isn't a role where AI is a feature bolted onto a product. The product is the AI: a proficiency engine that scores skills from real evidence, retrieval and agent systems that guide learning and career decisions, and data pipelines that turn organizational data into actionable skills intelligence. You'll design, build, evaluate, and ship these systems to production, where they directly influence real enterprise users' decisions about people and careers.

You'll work in a small engineering team that ships quickly, using AI tools (including AI coding assistants) as a standard part of how we build. You'll own real systems from day one—the code you write goes in front of live customers, which is exactly why it has to be rigorous.

What You'll Do

• Build LLM-powered features end to end on Azure: retrieval pipelines, agent workflows, prompt systems, and the APIs that serve them inside the product

• Own evaluation for AI features—define what "good" looks like, build evaluation harnesses, measure quality systematically, and iterate until the feature earns its place in production

• Develop skills inference and proficiency models that turn evidence (assessments, work artifacts, learning signals) into skill scores customers can trust

• Engineer data foundations in Microsoft Fabric: source, clean, and structure customer data from HR, learning, and job systems so AI pipelines can use it

• Take models to production: optimize for latency, cost, and reliability, and maintain the MLOps loop of versioning, monitoring, and retraining

• Deploy on Azure with clean APIs, containers, and CI/CD as standard practice

• Work daily with product and engineering peers to shape what gets built, explain tradeoffs clearly, and document decisions

• Practice responsible AI: fairness, transparency, and explainability are engineering requirements, not afterthoughts

What We're Looking For

• Strong Python and the modern AI stack. You're fluent with PyTorch or TensorFlow, Hugging Face, scikit-learn, Pandas, and NumPy. You write code other engineers want to inherit.

• Hands-on LLM application experience. You've built real systems with RAG, agents, tool calling, and structured prompting—including the vector search and retrieval infrastructure behind them.

• Experience with the Microsoft AI stack. OneGuru is built on Azure AI Foundry, Azure OpenAI, the Microsoft Agent Framework, and Microsoft Fabric. Direct experience here is a real advantage. Deep experience with equivalent tools plus appetite to go deep on ours also works.

• An evaluation mindset. You don't call an AI feature done because the demo worked. You build evals, test systematically, and use metrics to decide what ships.

• MLOps and deployment comfort. Docker, Kubernetes, CI/CD, and operating models on Azure are part of your toolkit.

• Software engineering fundamentals. Git, testing, API design, and the debugging patience to work through problems in large datasets.

• An AI-native way of working. You use AI coding tools daily as a core part of your craft, not an occasional helper.

• Clear communication. You can explain technical tradeoffs to non-technical stakeholders in plain language, in writing and in person.

• A track record of shipping. Years matter less than evidence—at least one AI or ML system you built that ran in production for real users.

Nice to Have

• Experience fine-tuning or adapting open-weight models, and depth in classical ML and NLP beyond LLM APIs

• A grounding in statistics and optimization

• Prior work in HR technology, learning, talent, or domains where model output directly affects decisions about people

What You'll Gain

• Ownership of the AI core of a product already in the hands of enterprise users

• A frontier tech stack and an AI-native team with freedom to adopt the best tools

• Real domain expertise in skills and talent, learned from live customers

• A clear path toward senior and lead AI engineering as the team scales

If you've built an AI system live in front of real users, evaluate rigorously before shipping, use AI tools as a core part of how you code, care about getting the technical details right because real decisions about people's careers depend on it, and are based in Noida and able to work onsite full-time, we want to hear from you! Apply now.