Abhay Jani
Embedded & Systems Engineer · Robotics & Computer Vision
C++, Python · FIRST Robotics · Axon R&D · USACO Platinum
I build autonomous and sensor-driven systems — from competition robots with real-time control and vision pipelines to industry R&D on embedded hardware validation. Based in San Jose, California.
I lead programming for FRC Team 2854 at Evergreen Valley High School, intern on ML and computer vision at Axon, and study engineering through dual enrollment at Evergreen Valley College.
Featured Work
FRC Team 2854 — The Prototypes
Lead autonomous programming and sensor integration for a 40+ member competition robotics team.
- Engineering Inspiration Award — FIRST Championship qualifier (2024)
- Autonomous routines with vision targeting, odometry, and PID motion profiling via WPILib
- Java/C++ robot code, sensor fusion, and field-tested control loops under competition constraints
Skills & Systems
Languages
Domains
Tools & Platforms
Signals
Projects
Problem: Competition robots must sense, decide, and act in under 15 seconds with unreliable field conditions and strict rules.
My role: Lead software for autonomous modes, sensor integration, and control tuning across FRC and FTC teams.
- Engineering Inspiration Award at FIRST Championship (2024)
- Shipped season robot code with autonomous scoring, vision alignment, and odometry-based navigation
- Led FTC 18715 as captain; coordinated mechanical, electrical, and software integration across subteams
- Mentored new programmers on Git workflows, sensor calibration, and competition debugging
Problem: Communities lack visibility into how data center growth stresses local electricity, water, and emissions limits.
My role: Built Python ingestion pipelines, geospatial rollup engine, and on-demand advocacy brief generation.
- 4,500+ mapped data center facilities nationwide
- 656 counties with rolled-up burden metrics
- 2,500+ advocacy brief downloads
- 8,000+ Kaggle dataset downloads
- Chose county-level geospatial rollups over per-facility queries for performance at scale
Problem: Multi-drone indoor navigation requires collision-free coordination in tight spaces with limited sensing.
My role: Built a custom Gymnasium environment on PyBullet, trained decentralized swarm policies with curriculum PPO, and benchmarked against a PID baseline.
- Procedural indoor environments with 16-ray lidar per drone
- 4-stage curriculum from single-drone navigation to 5-drone swarm coordination
- Interactive Three.js rollout replay with lidar visualization
- Extends classical FRC control experience into learning-based autonomy
Problem: Life Scouts struggle to find verified Eagle Scout service project opportunities and follow BSA workbook requirements.
My role: Sole builder of a Next.js platform with listings, org contacts, and an 8-phase completion guide.
- 21 curated project listings across the Bay Area
- 7 organizations with contact information
- 8-phase Eagle completion guide with checklists
Experience
- Developed and evaluated machine learning pipelines for computer vision applications on field hardware
- Prototyped embedded sensor integration workflows for bench testing and validation
- Collaborated with cross-functional R&D teams on electrical subsystem integration and test documentation
- Supported electrical engineering prototyping, wiring harness assembly, and bench-level hardware testing
- Coordinated intern team deliverables across mechanical, electrical, and software workstreams
- Documented test procedures and hardware validation results for R&D design reviews
Top competitive programming division — demonstrates algorithmic problem-solving under time constraints.
- Manage club operations, recruitment, and competition logistics for a multi-team robotics program
- Coordinate between FRC, FTC, and outreach subteams; see FRC 2854 project for technical detail
