Senior AI Engineer

Pepper; Mumbai; On-site; INR 420,000 - 700,000; Full time

Job summary

Pepper Content in Mumbai is seeking a Senior software engineer to own end-to-end outcomes for agentic AI systems. Design, build, and ship orchestration, memory, and LLM interfaces, collaborating across product and engineering to deliver rapid, user-impactful results.

You will work with multiple LLM providers and open-source models, implement tool calls, scale systems to production, and establish standards for reliable, cost-efficient AI at Pepper Content from Mumbai office.

Qualifications

• Built agentic AI systems in production.
• Hands-on with LLMs: prompt engineering, tools, retrieval.
• Experience with multiple LLM providers and open-source models.
• 6-10 years software engineering, fluent in Python.

Responsibilities

• Build and ship agentic AI systems end to end.
• Design and own the LLM interface layer: orchestration, tool calling, memory, evaluation.
• Work across many LLMs (proprietary and open-source).
• Shape product and technical direction with pragmatic focus.
• Set engineering standards for LLM and agent development.
• Iterate on models with business impact.

Skills

Python

NLP fundamentals

ML fundamentals

LLMs experience

Tools

OpenAI API

Open-source models

Model orchestration

Tool use

Job description

Company: Pepper Content Location: Mumbai (In-office) Employment Type: Full-time Seniority: Senior Experience: 6-10 years

About the Role

This is a strategic builder role for an engineer who wants to own outcomes, not tickets. You will design, build, and ship agentic AI systems end to end, moving fast, putting things in front of users, and scaling what works. You sit at the intersection of product and engineering, and you treat both as your job. This is an in-office role based in Mumbai.

We are looking for someone who has already built real agent systems: orchestration, tool use, memory, and the LLM interfaces that hold it all together. You are fluent across the model landscape, comfortable switching between providers and open-source models, and you know how to get the best out of each. You think in terms of what the product needs and ship at a pace that compounds.

What You Will Do

• Build and ship agentic AI systems end to end, from idea to production, at a fast and deliberate pace
• Design and own the LLM interface layer: orchestration, tool calling, memory, evaluation, and the glue that makes agents reliable
• Work fluently across many LLMs (proprietary and open-source), choosing and routing the right model for each job
• Shape product and technical direction together, deciding not just how to build but what is worth building
• Set the engineering standards and best practices for LLM and agent application development
• Iterate on models and systems with a sharp focus on real business and user impact
• Scale systems from first working version to production-grade reliability, performance, and cost efficiency
• Stay ahead of the frontier and bring the best new models, tools, and techniques into the product quickly

What We Are Looking For

• Proven experience building agentic systems in production, not just demos or prototypes
• Deep, hands-on experience with LLMs: prompt engineering, fine-tuning, tool use, retrieval, and evaluation
• Fluency across multiple LLM providers and open-source models, with strong judgment on tradeoffs
• 6-10 years of software engineering experience, strong fundamentals, fluent in Python, and comfortable owning systems end to end
• Solid grasp of NLP, ML, and LLM fundamentals
• Experience building, deploying, and scaling ML and LLM pipelines in production
• A genuine product sense: you can hold both the user problem and the engineering reality at once
• A strong bias for action, fast iteration, and continuous learning
• Deep interest in the latest open-source models, agent frameworks, and AI advancements

Who Will Thrive Here

You will do your best work here if you like ambiguity, move quickly without breaking trust, and care as much about shipping the right thing as shipping it fast. You want real ownership over what gets built and how, and you measure yourself by what reaches users and scales.