Armaan Sengupta
Skills
Software: Python, C, C++, Rust, Git, CMake, Bash, GBD, Protobuf | ROS2, FreeRTOS, Pytest, NumPy
Embedded Systems: STM32, NXP, Ambiq, ESP32, AVR | I2C, SPI, CAN, gRPC, UART, JTAG, BLE, MIPI DSI
Tools: KiCad, Altium, MATLAB, Simulink/Simscape, SimulationX, SolidWorks, Inventor, Fusion 360
Robotics: Controller Design, Forward/Inverse Kinematics, Trajectory Planning, State Estimation
Work Experience
Implant Firmware & Embedded Systems Intern
California, USA
Neuralink
Jan 2026 – August 2026
-
Led firmware development for 1st gen neural stimulation implant, implementing interrupt driven, DMA fed, high throughput code coordinating waveforms and sampling for 3k+ electrodes (Rust & C)
-
Architected the full app to ASIC stack for Blindsight, our human visual prosthesis product
-
Shipped code reducing registration time from 8m to 15s per trial enabling delivery of ~40M pulses per electrode in 13 weeks; making the safety study feasible and hitting FDA submission deadline (Rust)
-
Brought up impedance & calibration, enabling self test of chips, unblocking production line
Vehicle Firmware Body Controls Intern
California, USA
Tesla
May 2025 – August 2025
-
Brought up 20 exterior lights & related ADCs on Semitruck from hardware to app layer (Altium & C)
-
Fixed Cybertruck (CT) bug causing left/right headlamp stepper motors to calibrate out of sync (C)
-
Refined auto high-beam latching logic to engage only at night, improving user experience (C)
-
Developed current-sense self-tests for Cybercab lights, preventing faulty units passing production (C)
-
Fixed error in fault logging impacting Model 3/Y & CT, improving retroactive failure analysis (C)
-
Identified issue with short ADC sample time causing self-tests to false pass on Cybercab (Altium & C)
-
Wrote SIL tests verifying lighting functionality across models preventing regressions (Python: Pytest)
Embedded Systems & Firmware Engineering Intern
Indiana, USA
Ixana
August 2024 – December 2024
-
Designed firmware to implement video transmission from smart glasses to smart watch improving frame rate from 4 fps to 13 fps and reducing power consumption by 8% (C++)
-
Implemented automatic gain control for custom RF chip improving packet reception rates by 45%
-
Developed amoled display driver using MIPI DSI protocol enabling reliable video rendering (C++)
-
Developed library for Rigol Arb signal generator to facilitate HIL testing of custom silicon (Python)
-
Debugged SPI & I2C signal timing and encoding issues between FPGA & MCU with logic analyzer
Data Science & Controls R&D Engineering Intern
Ontario, Canada
Magna
January 2024 – April 2024
-
Designed an anomaly detection algorithm using strain gauge data from a hydraulic press to identify cracked parts and automatically filter them from the production line with a 100% accuracy during 150 part burst run (Python)
-
Develop an optimized data aggregation algorithm for a PLC using multithreading with custom semaphores and data queues to transmit data to a PostgreSQL DB, improving the data collection rate from 142 Hz to 500 Hz
Robotics and Embedded Systems Intern
Ontario, Canada
Fyelabs
May 2023 – September 2023
-
Designed a LiDAR dust filtering algorithm, enhancing object localization in harsh environments (C++)
-
Developed PCB board for a custom HMI panel enabling driver-system communication (EasyEDA)
-
Developed UART driver for LiDAR module to enable data logging and live HIL debugging (Python) & (C)
Design Team & Research
Research Assistant
Robotic Interaction, Perception and Learning Lab
September 2025 – January 2026
During my time at the RIPL Lab, I worked on a research project focused on contact localization in mobile robots; using joint torque sensing to identify where external contact occurs along the robot’s structure during operation. To support this, I developed a MuJoCo-based simulation bridge that allows our motion-planning missions to run either on the real robot or in a high-fidelity physics simulator. This integration replaced the previous RViz/Gazebo workflow and enabled realistic external force interactions, letting us inject contact events in simulation and read the resulting joint torque responses. I packaged the Franka Emika Panda MuJoCo model, implemented a ROS2 bridge node to synchronize joint states and trajectory commands, and extended joint limit handling to account for simulation precision. This work now provides the team with a testable, physics-grounded environment for advancing contact-based control and safe interaction strategies for robot manipulation.
Embedded Systems, Mechanical Design, & Controls
Waterloo Rocketry
September 2023– December 2024
I worked across the full stack on Waterloo Rocketry’s active airbrakes subsystem resulting in the team coming 1500ft closer to our target apogee at competition in 2024. My work covered designing a custom linear cam-and-slot mechanism driven by a single servo, running CFD simulations to map drag and maintain stability, and co-writing the STM32 FreeRTOS flight firmware for state estimation and trajectory prediction. The following year, I designed a 4-layer central avionics PCB that bumped our power efficiency to 95% and handled telemetry across multiple IMUs to drive our next project, a roll-control canards system.
Education
University of Waterloo
September 2022 - April 2027
Bachelors of Mechatronics Engineering
GPA: 90% (Dean’s Honors Role)