SWQA Triage Engineer

Role Overview

IndiVillage is hiring an SWQA Triage Engineer to support an active autonomous vehicle (AV) program. In this role, you will investigate, debug, and triage software quality issues across the full AV stack - perception, planning, and control - working directly with in-car and simulation logs, and channeling validated findings to Dev module leads and cross-matrix architects.

This is a domain-anchored engineering role distinct from data labelling: you will be analyzing and interpreting data and model outputs, not annotating raw data.

Key Responsibilities

• Root-cause software quality issues across perception, planning, and control; debug real-time sensor, compute-platform, and AI model failures
• Analyze sensor logs and simulation outputs at scale to identify failure patterns and trends
• Measure and track software stack and end-to-end DNN algorithm performance over time
• Review feature requirements, specifications, and design documents; communicate findings clearly to stakeholders
• Build debugging tools and dashboards; automate log analysis and anomaly detection; improve metrics, tracing, and monitoring

Required Qualifications

• B.Tech or equivalent degree in CS, CE, Robotics, ECE, or EEE
• 3+ years of hands-on testing and debugging experience in autonomous systems or computer-vision projects
• Python proficiency, including the ability to write logical scripts and code from scratch
• ADAS verification and validation experience with HIL/SIL
• Working knowledge of Linux and/or QNX
• Understanding of ISO 26262 functional safety as applied to autonomous systems

Preferred Qualifications

• On-road or test-track experience
• ML or self-driving certification
• Hands-on experience with driving simulators such as IPG CarMaker or TASS PreScan

Additional Attributes

• Strong analytical and structured problem-solving ability
• Excellent collaboration and teamwork skills
• Clear, proactive communication in English, including reporting to non-technical stakeholders
• Comfort operating independently in a fast-iteration, high-ambiguity environment