Applied AI Research Engineer

About the role

You define the quality bar for AI-powered features by designing eval rubrics, test plans, and rollout criteria that are measurable and enforced in production. You build and extend production code while setting up monitoring and tests that catch regressions before users encounter them, owning features end to end from problem framing through rollout and iteration. You debug failures across the full stack including data, infrastructure, model logic, and prompt design, then harden systems with the insights you gain. You design and implement systems such as retrieval pipelines, agents, or hybrid patterns based on what the problem actually needs rather than following trends. You work across functions collaborating with product, infrastructure, and engineering teams to ship features that are reliable and stick in production. You translate ambiguous customer feedback problems into concrete AI system requirements and success metrics that guide the entire lifecycle of an AI feature. You carry the load when things break post-launch, driving investigations, fixes, and improvements that keep the system reliable for customers.

Key facts

What you'll do

Design eval rubrics, test plans, and rollout criteria that define what good looks like for AI-powered features and ensure they are measurable and enforced.

Write and extend production code, set up monitoring, and add tests that catch regressions before users do, keeping systems reliable in real-world conditions.

Own features end to end from problem framing to modeling, from system design to rollout and iteration, taking responsibility for outcomes across the lifecycle.

Debug failures across the stack including data quality, infrastructure behavior, model performance, and prompt logic, then harden systems with lessons learned.

Design and implement systems such as retrieval pipelines, agents, or hybrid patterns based on the actual requirements of the problem rather than fashionable approaches.

Collaborate with product, infrastructure, and engineers to ship features that actually stick, aligning on goals, constraints, and tradeoffs.

Translate ambiguous customer feedback problems into concrete AI system requirements and success metrics that guide experiments and evaluations.

Run experiments and analyses to measure the impact of AI features on business outcomes, using insights to prioritize further work and improvements.

Contribute to the broader research agenda by documenting findings, sharing learnings, and proposing new directions that improve how feedback is understood and acted on.

Partner with cross-functional stakeholders to gather context, define priorities, and ensure that AI features support the mission of unlocking the voice of the customer for every product team on the planet.

Requirements

You have built and shipped AI systems before and carried the load when things broke post-launch, demonstrating ownership and reliability under pressure.

You have strong research instincts, including the ability to define what working means and design evaluations that reflect real-world usage and business value.

You are comfortable writing and extending production code, with the maturity to set up monitoring, tests, and guardrails that protect users from regressions.

You understand retrieval, agents, and hybrid patterns well enough to choose or combine them based on the problem, not the other way around.

You communicate clearly and work effectively across functions, aligning with product, infrastructure, and engineering teams to keep features moving forward.

You are rigorous in measurement, insisting on measurable criteria for success and enforcing them through experiments and analysis.

You are comfortable operating in a fast-moving environment where priorities shift and problems are ambiguous, requiring judgment and ownership to resolve.

You care deeply about customer feedback and are motivated by the mission of making the voice of the customer actionable for product teams everywhere.

Nice to have

Experience working with customer feedback, surveys, reviews, or support tickets in real products.

Familiarity with large language models, prompt engineering, and evaluation methods for generative AI systems.

Experience building systems that involve retrieval, reasoning, and agentic behavior, and understanding their tradeoffs.

Practical notes

This is a full-time role based in Bengaluru.

There are no travel, visa, or application deadline requirements specified in the source.