AI Platform Evaluation Software Engineer - Autonomous Vehicles

Role Overview 

As a member of the  core AV software reliability team , you will be responsible for ensuring  safe, stable, and scalable Autonomous Vehicle (AV) software releases  by turning failures into actionable engineering insights at scale. 

This is a  software-first, platform-focused  role. You will work primarily across the  AV platform software stack  (frameworks, runtime, services, orchestration, data pipelines) and its interaction with  vehicle hardware and compute —not cloud infrastructure or hardware design. 

The mission of this role is to: 

• Improve  learning velocity from failures  

• Reduce  reliability escapes  

• Increase  confidence in production readiness  

You will do this through  intelligent triage, deep software debugging, and AI-assisted failure analysis  across  simulation, CI, HIL, SIL, and on-road environments , ensuring that failures are: 

• Correctly detected and interpreted 

• Consistently categorized and de-duplicated 

• Rapidly mapped to the right owners and solution space 

You will collaborate closely with  AV software engineers, ML engineers, systems engineers, test platform owners, and release/safety stakeholders  to ensure reliability signals directly influence  engineering priorities  and  release decisions . 

If you are passionate about  software reliability, failure analysis, and building AI-driven systems  that help organizations learn faster from complex ML-based AV software, this role is for you. 

Key Responsibilities 

• Own the AV software reliability triage framework  for the on-vehicle / AV platform stack, defining how failures from  simulation, CI, HIL/SIL, and on-road validation  are detected, grouped, and escalated into actionable tickets and insights. 

• Perform deep debugging and root-cause analysis  across: 

• AV platform and framework code 

• Perception / planning / control software integrations 

• ML pipelines and model rollouts 

• Vehicle compute and hardware interfaces (sensors, ECUs, networks) 
  Connecting failure symptoms (logs, time-series, traces) to clear solution paths and corrective actions. 

• Design and evolve automated triage mechanisms and reliability taxonomies  that: 

• Improve regression detection and signal-to-noise ratio 

• Identify flaky tests and intermittent platform issues 

• Enable consistent failure classification across teams and releases 

• Build and govern reliability data pipelines  for AV software: 

• Ingest logs, metrics, traces, and test results across CI, simulation, and vehicle runs 

• Compute stability trends, recurrence patterns, and systemic risks 

• Provide dashboards and views that support day-to-day triage and release readiness reviews 

• Apply AI / ML to reliability triage , for example: 

• Clustering and de-duplicating failures across large-scale test runs 

• Learning-based suggestions for ownership, component mapping, and likely root causes 

• LLM-powered summaries of complex failure scenarios for engineers and leadership 

• Translate reliability findings into decision-grade communication : 

• Influence prioritization of bugs vs. technical debt vs. feature work 

• Provide input to go/no-go decisions and safety/release governance 

• Create clear narratives and visuals for engineering, safety, and leadership audiences 

Required Qualifications 

• Strong proficiency in  Python  for automation, log analysis, data processing, and reliability tooling. 
   

• Proficiency in  SQL  for querying reliability and test data, building views, and supporting dashboards. 

• Proven experience with  CI/CD systems  (e.g., GitHub Actions, Jenkins, GitLab CI or equivalent) used to run automated tests for complex software systems. 

• Hands-on experience implementing  ETL/ELT pipelines  for reliability, quality, or system health monitoring (e.g., ingesting logs/metrics from test runs, building reliability datasets). 

• Solid understanding of  software reliability engineering concepts , including: 

• Regression tracking 

• Flakiness detection and management 

• Failure classification and de-duplication 

• Release criteria and quality gates for software 

• Strong analytical and cross-stack debugging skills in  large-scale distributed or real-time software systems , ideally with  C++ and/or Python-based services . 

• Experience integrating  simulation, HIL/SIL, or system-level AV/robotics test signals  into automated analysis workflows. 

• Track record of effective cross-functional collaboration with  software engineering, QA, test platform, and systems teams . 

• Ability to  operate autonomously  in high-ambiguity, safety-critical environments, driving clarity and decisions from noisy data. 

• Excellent written and verbal communication skills for presenting  data-driven reliability insights  to engineers and technical leadership. 

• Bachelor’s, Master’s, or PhD  in Computer Science, Electrical/Computer Engineering, Robotics, or a related field — or  equivalent software-focused experience . 

Preferred Qualifications 

• Experience with  reliability governance in ML-based or AV systems , including: 

• Model rollout policies 

• Guardrails / kill-switches 

• Shadow or A/B validation strategies 

• Familiarity with  reliability methodologies  (e.g., FMEA, reliability growth analysis, MTBF trends) applied to software and integrated AV platforms. 

• Knowledge of  AV / ADAS software architectures , including: 

• Perception / localization / planning / control pipelines 
   

• On-vehicle platform software, middleware, and simulation-to-road validation loops 

• Experience building  reliability or analytics pipelines in cloud environments  (AWS, GCP, Azure) to support AV software validation at scale. 

• Familiarity with  observability and visualization tools  (e.g., Grafana, Superset, Power BI, or similar) for reliability dashboards and on-call / triage workflows. 

• Experience using  Jira, GitHub Projects , or similar tools to design reliability triage workflows, routing, SLAs, and dashboards. 

• Experience applying  AI / ML or LLMs  to: 

• Log analysis and anomaly detection 

• Failure clustering and root-cause suggestion 

• Automated summarization of complex test outcomes