Mukund Jeedigunta

Robotics software engineer specializing in end-to-end robotic software development: ROS2-based autonomous navigation and localization, AI-driven perception and voice interaction, and computer vision. Experienced across the full robotics stack, from embedded systems and PCB design to simulation-to-hardware deployment using Gazebo, Nav2, and custom ROS pipelines.

Autonomous Systems & AI Engineer (Senior Design Project) at PAWS Rover (2026-05 – Present)

Co-designing an autonomous outdoor rescue rover that uses computer vision and audio sensing to locate distressed stray cats and dogs, deliver food, and relay live GPS location and video to an operator.

• Own the simulation and navigation deliverable: built and validated a full autonomous mission cycle in Gazebo combining a lawnmower-pattern area search, A* path planning, and Vector Field Histogram (VFH) obstacle avoidance, driving a finite-state machine through Searching, Approaching, Delivering, and Returning.
• Implemented GPS waypoint navigation using Haversine bearing calculations and a proportional heading controller to convert bearing error into differential drive commands, engineered within a $492 total hardware budget (~10x cheaper than the nearest commercial equivalent).

Robotics Engineer at EchoNav (2026-06 – Present)

Built EchoNav, a modular ROS2-based robot that translates natural-language voice commands (e.g., "go to the kitchen," "return home," "stop") directly into autonomous navigation actions, integrating Whisper for speech recognition, an LLM (Llama 3 via Ollama) for intent interpretation, and Nav2 for navigation execution.

• Designed a multi-node ROS2 architecture (voice input → intent recognition → task planning → navigation execution) with topic-based communication, service calls, and action interfaces, enabling modular, independently testable components and validated end-to-end in Gazebo simulation on TurtleBot3.

Robotics Engineer at Autonomous Navigation Robot (2026-03 – 2026-04)

• Built a two-layer ROS navigation stack combining A* global path planning with DWA local obstacle avoidance, achieving collision-free navigation in 95% of dynamic obstacle scenarios across 50+ test runs.
• Engineered a "Home Service" logic to manage multi-goal pickup/delivery tasks with zero human intervention.

Robotics Engineer at Monte Carlo Localization (AMCL) (2026-02 – 2026-03)

• Implemented AMCL by fusing LiDAR and odometry data to achieve precise pose estimation in non-static scenes.
• Reduced pose estimation error from 18cm to 6cm by tuning particle count and adaptive resampling thresholds.

Co-Founder and Lead Developer at Prompt Kart INC. (2025-06 – Present)

Designed and built PromptKart, a SaaS platform delivering curated AI prompt packs to professionals, creators, and startups across 6 content verticals.

• Engineered the full-stack architecture (JavaScript, REST APIs, secure authentication) with content indexing and search/filter pipelines, reducing query latency by 45% in staging benchmarks.

Lead Student Engineer at UMass M5 Engineering Lab (2024-08 – 2025-01)

• Designed a custom PCB integrating a microcontroller-based embedded system with peripheral device communication.
• Optimized firmware execution flow under real-time constraints, improving system stability by 30%.

Bachelor of Science in Computer Engineering – University of Massachusetts Amherst (2026-12)

Nanodegree in Robotics Software Engineer – Udacity