Supratim Chatterjee — Senior Design Verification Engineer

verify

/ˈvɛr.ɪ.fʌɪ/ · verb

to establish the truth or correctness of something through rigorous, repeatable testing.
(in silicon) to prove a design behaves exactly as intended — across every corner, every cycle, every condition — before it becomes permanent.

01 — About

Precision is the whole job.

I work as a Senior Design Verification Engineer at NXP Semiconductors in Bangalore, verifying mixed-signal automotive ICs — LED controllers and battery-management systems — the kind of silicon that has to be right because it ends up in cars. Day to day that's SVA properties in Cadence JasperGold, coverage closure, protocol testbenches, Python/AI automation, and building custom MCP servers & autonomous agents that boost verification productivity 2–3×. Five successful tapeouts owned, bring-up to sign-off.

That verification mindset — assume nothing, test the boundaries, trust only what you can prove — bleeds into everything else. Before NXP I built audio-EQ and voice-detection ML at Texas Instruments and a tyre-inspection vision system at Tata Motors; on my own time I tinker with AI agents, classifiers, IoT alerts, and writing about ideas I can't stop chewing on.

BITS Pilani, Hyderabad alumnus (B.E. Electronics & Instrumentation, minor in Robotics & Automation) and currently pursuing an M.Tech in Artificial Intelligence at IIT Madras — happiest when I'm pulling apart a problem that sits just outside whatever I already know how to do.

Design Verification Signal Processing Machine Learning Digital Design Internet of Things

02 — Experience

How I got here

From a stair-climbing robot in a campus competition to owning sign-off on automotive silicon.

WorkApr 2024 — Present

Senior Design Verification Engineer

NXP Semiconductors · Bangalore

Engineered scalable AI verification systems by developing custom MCP servers, autonomous agents, and structured workflows integrated into internal tools. Verified mixed-signal automotive ICs for LED controllers and BMS, ran formal property & connectivity verification in JasperGold, and built Python/AI automation that boosted verification productivity by 2–3×. Presented Regression Run Time Optimization using Verisium Smart Run at Cadence Connect.

5 tapeouts ownedMCP servers & agentsJasperGold FPV + Connectivity2–3× productivity

WorkJul 2023 — Apr 2024

Design Verification Engineer

NXP Semiconductors · Bangalore

Developed and debugged verification environments using CAN, SPI, I2C, JTAG and UART protocols. Executed formal property and connectivity verification in Cadence JasperGold, authoring SVA properties and constraints, and drove coverage closure across critical IP blocks.

JasperGold FPV + ConnectivityCAN · SPI · I2C · JTAG · UART

PublicationJan 2024

A Novel Approach for Automated Speaker Equalization ↗

IEEE Xplore

A method that predicts speaker-equalization filter coefficients directly from the desired and measured frequency responses — replacing the manual tuning audio engineers do by hand.

Peer-reviewedSignal Processing

InternshipJan 2023 — Jun 2023

Software Engineering Intern — AI / ML

Texas Instruments · Bangalore

Built an automatic and dynamic (signal-level-aware) audio-EQ system that generates filters to hit a target frequency response — cutting tuning time and manual error — plus a voice-activity-detection model robust in low-SNR conditions and light enough to run on edge devices.

Audio EQ automationEdge-deployable VAD

EducationPresent

M.Tech — Artificial Intelligence

Indian Institute of Technology, Madras

Postgraduate specialisation in Artificial Intelligence at one of India's top-ranked engineering institutions, deepening foundations in machine learning, deep learning, and AI systems.

IIT MadrasArtificial Intelligence

EducationAug 2019 — Jul 2023

B.E. Electronics & Instrumentation

BITS Pilani, Hyderabad

Minor in Robotics & Automation. Coursework across Embedded System Design (graduate), Analog & Digital VLSI Design, FPGA-Based System Design, Microprocessors & Interfacing, Neural Networks & Fuzzy Logic, and Machine Learning.

8.93 / 10 GPAMinor: Robotics & Automation

InternshipMay 2022 — Jul 2022

AI Intern

TML Business Services (Tata Motors) · Pune

Built a computer-vision system to read key features off vehicle tyres for warranty claims — object detection to isolate the text regions, OCR to read them, and a generated QR code per tyre — deployed on-site on Raspberry Pi and Jetson Nano.

Vision + OCREdge: RPi · Jetson Nano

InternshipMay 2021 — Jul 2021

Summer Intern

CSIR — National Institute of Oceanography · Goa

Research internship at one of India's premier oceanographic institutes, gaining exposure to scientific computing and data-driven methods in an applied research environment.

ResearchGoa

CompetitionOct 2020

Flipkart Grid 2.0 — National Semifinalist

National-level robotics competition

Designed a stair-climbing warehouse robot able to carry a 3–5 kg payload across flat floors and standard staircases. Reached the national semifinals.

Semifinals

AwardOct 2019

CBSE Certificate of Merit — Physics

Central Board of Secondary Education

Awarded for placing in the top 0.1% of candidates in Physics in the Higher Secondary Board examinations across India.

Top 0.1% nationally

03 — Toolbox

The arsenal

The tools I reach for across the verification suite, the lab bench, and everything in between.

Languages 04

SystemVerilogPythonVerilogC++ / C

Frameworks 05

UVMPyTorchTensorFlowOpenCVMCP

Tools & Platforms 06

VmanagerSimvisionJasperGoldVerisium DebugMATLABVivado / Xilinx

Protocols 06

SPII2CCANTPLJTAGUART

04 — Domains

What I work on

D1

Verification

SVA, formal property & connectivity verification, and coverage closure on mixed-signal automotive ICs.

D2

Machine Learning

CNNs, computer vision, and signal-processing ML for imaging, audio EQ, and detection — TensorFlow & OpenCV.

D3

IoT & Embedded

Connected alert systems and embedded logic — from crypto price pings to billing hardware.

D4

Writing

Explaining hard ideas in plain language across tech and cricket on my blog and Medium.

05 — Selected Work

Things I've built

A sample of projects. Everything links straight to the source.

↗

Python · IoTTelegram

Cryptostock

An IoT-based alert system that messages you on Telegram the moment a chosen cryptocurrency crosses your target price threshold.

IoTAutomationAPIs

↗

Jupyter · Deep Learning~98% acc.

Disease Detector

A CNN that classifies chest X-rays as COVID vs. non-COVID, reaching ~98% accuracy on the validation set. Built for the Remote Sensing & Image Processing course.

CNNMedical ImagingImage Processing

↗

Jupyter · FundamentalsBITS F312

Neural Networks & Fuzzy Logic

Core ML and fuzzy-logic algorithms implemented from scratch in Python — built for the BITS F312 course, no black boxes.

From ScratchNumPy

↗

Verilog · FPGAZedboard Zynq

Automatic Billing System

An FPGA-based shopping-cart billing system that tallies multiple items as they're added and zeroes out for the next transaction. Implemented on the Zedboard Zynq evaluation kit.

VerilogFPGADigital Design

↗

Jupyter · SimulationAug 2024

Human Brain Storage Simulator

An implementation of a published memory-storage model, simulating how the human brain encodes and retains information.

Cognitive ModelingResearch

↗

Jupyter · Computer Vision2026

Wafer Map Defect Detection

A model that classifies defect patterns on semiconductor wafer maps to speed up fab yield analysis.

SemiconductorsImage Classification

All projects on GitHub ↗

06 — Publications

Peer-reviewed work

Research published alongside the engineering.

Jan 2024

IEEE Xplore · Peer-reviewed

A Novel Approach for Automated Speaker Equalization

Predicts speaker-equalization filter coefficients directly from the desired and measured frequency responses — automating what audio engineers tune by hand.

↗

07 — Writing