# Narendar Sriram > Physical AI @ OSARO | MRSD ‘23, CMU | Physical AI, VLA, Data - Robotics Location: San Francisco, California, United States Profile: https://flows.cv/narendar Winner - Best use of Open AI, Hack agents hackathon | Winner - Best use of Groq, AI4Good Hackathon. 6 years of industry experience. 3 years in Robotics. I build vision-language models and learning-based robotic systems that work in production. Presently, I am an ML and Software Engineer at Osaro developing VLAs for industrial robots (first in the industry). My work spans the full stack - from data pipelines and model training (Gr00t, Pi0, PyTorch) for industrial manipulators to motion planning, fleet orchestration, behavior planning, robot navigation, state machines and real-time execution. I've built end-to-end learning pipelines on FANUC manipulator that hit 92% policy accuracy, built a SLAM based natural navigation system at KUKA, architected deterministic systems for heterogeneous robot fleets and solved multi-agent planning problems at scale using Hybrid A* and Graph search. Before specializing in learning-based manipulation, I led navigation teams that worked closely with GM and Panasonic and deployed autonomous systems in live warehouses. Started as a software engineer at Oracle optimizing large-scale enterprise systems before moving into robotics. Trained in Carnegie Mellon, Robotics Institute. Six years of building systems that learn from data, handle edge cases, and run reliably when it matters. Portfolio: https://narendar-sriram-portfolio.netlify.app/ ## Work Experience ### Sr Robotics Software Engineer @ OSARO Jan 2024 – Present | San Francisco Bay Area Senior Robotics Software Engineer at OSARO. Architected and developed a scalable end-to-end VLA data flywheel system with a team of 4 engineers for FANUC lr-mate using Python, GR00t N1.5 and PyTorch, trained on real + simulated dataset improving overall policy accuracy to 92%. Built large scale data ingestion, labeling pipeline for visual, spatial and sensor data, enabling systematic evaluation of model performance, and behavioral regressions across experiments. Integrated real-time chunking safe trajectories, ensuring reliable execution on FANUC lrmate. Developed a mission-critical Rust gRPC fleet-coordination business logic software orchestration pipeline for a complex heterogeneous robot fleet providing live production debugging, conflict resolution, fault recovery and zero-downtime handoffs in production. Successfully developed the end-to-end robot reachability software tool using PyQT, Open3D, Rust and Bevy to load robot and world meshes, analyze reachability of all the points using kinematics. ### Senior Robotics Software Engineer @ KUKA Jan 2023 – Jan 2024 | Sterling Heights, Michigan, United States Working on Natural Navigation, Traffic Control and Path planning for Kuka AGVs ### Research Assistant @ Robotics Institute, CMU Jan 2022 – Jan 2023 | Pittsburgh, Pennsylvania, United States Working on agent to agent collision avoidance techniques in Multi Agent Path planning using control barrier functions. ### Robotics Software Intern, Motion Planning @ Path Robotics Jan 2022 – Jan 2022 | Columbus, Ohio, United States Working on solving edge cases for constrained motion planning in a weld like scenario. I used state of the art optimisers such as TrajOpt and STOMP on seed plans constructed from RRT* and used it for moving frames in welding. ### Research Engineer in Mobile Robotics @ Addverb Jan 2020 – Jan 2021 | Noida, Uttar Pradesh, India Addverb Technologies | Research Engineer in Mobile Robotics Delhi NCR, India | Aug 2020 - Jul 2021 • Developed task and scheduling algorithms for a coordinating fleet of ~300 robot agents using Heat Map, A* Planner & traffic handling solutions for warehouse sorting; carried out simulation testing using OpenCV & increased the throughput by 100% • Developed a Multi Agent Path Planning system as the Lead FMS architect using occupancy grid mapping for Zippy (a pick and place sorting Robot) from ground up and deployed at Panasonic, India. Features included: o Server client architecture, Back End for UI and communication with PLCs using API calls featuring no latency or loss of data. o Robust, database engulfing Task Planner using MySQL and C++; Developed FIFO based reservation tableand did rugged testing on robots for docking, obstacle avoidance & automatic charging. • Engineered a goal mapping Multi Agent Path Planner and an exhaustive State Machine using C++ for a naturally navigating pick and place robot with a payload of 100kg; developed system successfully communicated with client robots, performed tasks without interference & loss of throughput, managed traffic & sent data to back end via APIs (deployed at JK Tyres, India). ### Application Engineer @ Oracle Jan 2017 – Jan 2020 | Hyderabad, Telangana • Developed user interface (UI) that include new software features and customer-facing UI changes as a part of Cost Accounting Product within Oracle India’s Supply Chain Management. • Spearheaded the development team that optimized SQL queries in Receipt Accounting Management Software through efficient data structure design and identifying performance bottlenecks, bad Java methods and DB object failures. This helped reduce the processing cost by 30%. • Independently designed the modules used in the automation of Cost Adjustment flow as an Automation Design Engineer and worked with a team of developers in implementing it with zero errors. ### Control Engineer @ Reliance Industries Limited Jan 2016 – Jan 2016 | Gujarat, India • Worked on the effect of PID parameters as a part of the reputed summer internship program of Reliance Industries Limited and worked as a Performance Control Engineer in the Manufacturing division in Honeywell, Yokogawa and Emerson DCS controllers at Dahej Manufacturing Division (DMD, RIL); studied the working of Control Valves, Relays and Shut off Valves in detail. • Analyzed the working of Human Machine Interaction (HMI) System and the effect of P, I and D parameters on them and improved the process efficiency by 15%. ## Education ### Masters in Robotic Systems Development Carnegie Mellon University ### Bachelor of Technology - B Tech in Instrumentation and Control Engineering National Institute of Technology, Tiruchirappalli ### Masters in Mechatronics, Robotics, and Automation Engineering Carnegie Mellon University ### Masters in Mechatronics, Robotics, and Automation Engineering Carnegie Mellon University ## Contact & Social - LinkedIn: https://linkedin.com/in/narendar-sriram - Website: https://narendar-sriram-portfolio.netlify.app/ --- Source: https://flows.cv/narendar JSON Resume: https://flows.cv/narendar/resume.json Last updated: 2026-04-01