return to home
Reach Out
Résumé

David Nash

Robotics & AI/ML Engineer
Email
Summary

Software engineer and mechatronics-trained roboticist transitioning from four years building large-scale distributed systems at AWS into remote robotics consulting, with a focus on AI/ML simulation and control. Currently building applied expertise in MuJoCo physics simulation, reinforcement learning, and lightweight on-device AI (state-space models, Liquid Neural Networks) for robotic arm control. Brings a rare combination: formal mechatronics and embedded systems training, four years of production-scale ML and distributed-systems engineering at AWS, and a self-built, launched full-stack product — well suited for remote engagements in robotics simulation, AI/ML integration, and software architecture.

Skills
Robotics & AI/ML Simulation: MuJoCo simulation, applied reinforcement learning, inverse kinematics, state-space models (Mamba), Liquid Neural Networks for on-device control, computer vision (Python), embedded systems (C, Assembly, Verilog, Arduino, LabVIEW), digital logic and circuit fundamentals, working knowledge of ROS
AI/ML & Data Engineering: Practical ML pipeline design and deployment, ensemble modeling, ML data preparation, Apache Spark, Apache Parquet, ETL and data warehousing, large-scale distributed data processing (100M+ records/day)
Cloud & Backend Engineering: AWS (Lambda, Fargate, S3, AppConfig, EventBridge, Athena, SageMaker, Glue), serverless and cloud-native architecture, distributed systems, event-driven architecture, Java, Python, TypeScript/JavaScript, SQL, Bash, C#, Groovy, API design, Docker, infrastructure as code, CI/CD pipelines, testing (unit/integration/functional, Spock)
Full-Stack & Product Development: Svelte, TypeScript, Vite, Vitest, Tailwind CSS, Turso DB, Drizzle ORM, Supabase, PostGIS, Cloudflare (Workers KV, WAF, Turnstile), LLM/RAG integration, Unity (VR development)
Experience
Independent — Robotics & AI/ML Consulting (in progress)
2026–Present
  • Transitioning from software engineering into robotics, building applied expertise in physics simulation and embedded AI control for remote consulting work
  • Building hands-on proficiency in MuJoCo for robotic simulation and applied reinforcement learning
  • Prototyping on-device control approaches for robotic arms using state-space models (Mamba) and Liquid Neural Networks, targeting lightweight, low-latency inference on embedded hardware
  • Developing original coursework on MuJoCo, applied reinforcement learning for robotics, and foundational topics (inverse kinematics, embedded systems) to teach others — demonstrating both technical depth and the ability to communicate complex concepts clearly
Founder & Full-Stack Engineer — StayMapper
2025–2026
  • Designed and built a full-stack travel web application end-to-end, solo, using Svelte 5, TypeScript, Tailwind CSS, Supabase, and Cloudflare
  • Built an interactive map component (MapLibre) rendering 400,000+ points of interest via optimized PostGIS geospatial queries
  • Reduced load times by orders of magnitude with a custom Cloudflare KV caching layer
  • Integrated an LLM-powered chatbot with retrieval-augmented generation (RAG) over proprietary city-guide content to answer customer travel questions
  • Built data pipelines and SOPs to scale content production to 60+ fully detailed city guides, updated weekly
  • Implemented bot and fraud mitigation (Cloudflare Turnstile, WAF rules, honeypots, behavioral analysis) and integrated third-party APIs (Booking.com, Geoapify, Serper.dev, LemonSqueezy) for data and revenue
  • Took the product from architecture through launch and monetization as a solo engineer
Software Engineer II — Amazon Web Services (AWS Billing)
2020–2024
Hired March 2020 as SDE I (new-grad hire); promoted to SDE II, December 2021
  • Designed and built a scalable ML pipeline (AWS Lambda, S3, Athena, SageMaker, Glue) that parsed, enriched, and published hundreds of millions of billing files daily; trained an ensemble model to predict per-account compute time and memory needs at a 90% confidence interval for VIP accounts
  • Architected a distributed computing system (Billing Entity Mapper) that processed 16M+ accounts per hour on AWS Lambda, Fargate, and S3 to determine correct billing logic per account, cutting per-batch latency by 50% and eliminating computational waste
  • Served as technical SME for "Savings Plans on SDK" across engineering, product, and customer-facing teams: delivered 40–90% performance gains, a 10x increase in scaling capacity (4M → 40M alternatives), and raised test coverage from 75% to 95%
  • Built a serverless account-management system (CALM) that automatically identified and rebalanced 850+ of 17M billing accounts onto more performant infrastructure, preventing 50+ Sev-2 incidents
  • Led a cross-team billing platform change spanning 10+ packages and 50+ classes owned by different teams, coordinating with downstream stakeholders on schema changes to enable new international markets and currencies
  • Mentored new engineers during onboarding and authored reusable technical documentation templates later adopted by other teams
VR Specialist — Undergraduate Researcher, Maestro Labs, Texas A&M University
2017–2019
  • Applied computer vision (Python) to build robotic systems for strawberry-picking, line-following, and shape recognition, and to support mixed-reality capture workflows
  • Developed Unity-based virtual reality applications and simulations as part of ongoing lab research
VR Specialist Intern — Air Force Research Laboratory, Dayton, OH
Summer 2018
  • Built a peer-to-peer networked virtual reality scene in support of defense research applications of immersive simulation
Education
B.S., Mechatronics Engineering (minor in Embedded Systems)
2015–2019
Texas A&M University
  • Combined electrical and mechanical engineering coursework focused on robotics: control systems, inverse kinematics, digital and analog electronics, structural analysis, thermodynamics, fluid dynamics
  • Embedded systems minor: Assembly, C, LabVIEW, Verilog, Arduino, digital logic design