Tier 1: Advanced Industry CAD Automation & AI-Assisted Silicon (2023 – Present)

Showcasing cutting-edge agentic CAD toolchains, autonomous RISC-V FPGA tapeouts, and full-stack systems engineering.

Internal CAD Automation

Industry CAD Development | 2023 – Present

Architected internal CAD automation workflows leveraging Python and TCL. Engineered custom scripts and automation hooks to accelerate analog circuit sizing, formal equivalence verification, and DRC/LVS routing iterations across advanced semiconductor process nodes.

Efficiency Custom Scripting
Architecture Custom Skills & Hooks
Python Automation TCL Scripting EDA Workflows Automation Hooks Analog CAD

Antigravity_RISCV_64bit_Processor

Custom Architecture & Agentic Framework | 2024

Designed a custom 1 GHz 64-bit RISC-V processor accompanied by an Agentic Framework. Features AI agents that monitor, test, and improve the RTL design while reporting real-time metrics to a beautiful dashboard.

Target Architecture 64-Bit RISC-V Core
Framework Real-time Dashboard & AI
RISC-V 64-bit Agentic Framework RTL Dashboard

System-Level Mixed-Signal Modeling & Formal LEC

Enterprise ASIC Methodologies | 2023 – Present

Engineered system-level analog and mixed-signal verification environments leveraging XMODEL and MODELZEN by Scientific Analog. Executed rigorous equivalence checking and static property verification using Synopsys ESP and Cadence Conformal LEC across multi-million gate ASIC tapeouts.

Analog Modeling XMODEL / MODELZEN
Formal Verification Synopsys ESP / Conformal
XMODEL MODELZEN Synopsys ESP Conformal LEC Mixed-Signal

Tier 2: University & Research Foundations (NITK 2020 – 2023)

Undergraduate honors thesis, academic tapeouts, and foundational hardware engineering research at National Institute of Technology Karnataka.

RISC-V Microcontroller using OpenLane

Under Prof. Ramesh Kini, NITK | Jan 2022 – Apr 2022

Configured open-source RISC-V core PICORV32 with custom co-processors and modified the OpenLane flow for manufacturing GDSII files using SkyWater 130nm PDK. Performed extensive design space exploration across area and delay optimization.

Max Frequency 250 MHz Clock
Chip Area 0.49 – 0.70 mm²
Verilog OpenLane SkyWater 130nm RISC-V Core

Configurable Framework for CNN on Hardware

Under Prof. Sumam David, NITK | Jan 2023 – Apr 2023

Designed scalable memory architectures to efficiently utilize onboard FPGA BRAM to load input parameters of CNNs of any kernel size and stride length. Exploited pixel re-usability during convolution to drastically reduce LUT utilization.

Optimization Low LUT / BRAM
Scalability Any Kernel / Stride
FPGA Design Hardware Accelerator CNN Architecture Verilog

MIT Beta-ISA Pipelined Microprocessor

Under Prof. MS Bhat, NITK | Jan 2023 – Apr 2023

Architected a full 5-stage pipelined microprocessor (Instruction Fetch, Decode, ALU, Memory, Write-Back) based on MIT OCW Beta-ISA. Built a custom GUI-based assembler in Python to translate assembly code to machine code.

Architecture 5-Stage Pipeline
Toolchain Python GUI Assembler
CPU Architecture Beta-ISA Verilog Python Assembler

Security Module for Microprocessor

October 2020 – March 2021

Designed a 32-bit RISC processor in Verilog with its own ISA. Integrated a 128-bit AES encryption block to encrypt data/instructions in memory, thwarting hardware probing attacks. Developed a Python assembler that encrypts machine code on the fly.

Security Engine 128-bit AES Hardware
Protection Anti-Probing Memory
Hardware Security AES Encryption 32-bit RISC Verilog

4-bit ALU using Magic VLSI

Under Prof. Ramesh Kini, NITK | Nov 2021

Implemented a 4-bit ALU capable of Addition, AND, and OR operations using Pharosc vcslib 130nm technology standard cells. Wrote an automated bash testbench script that generates random test cases and verifies layout outputs.

Standard Cells Pharosc 130nm
Layout Area 169,061 sq. µm
Magic VLSI Custom Layout Bash Automation Python

Live Object Detection Using Raspberry Pi

Under Prof. Sumam David, NITK | Jan 2021 – Mar 2021

Deployed MobileNet-SSD V2 object detection model on Raspberry Pi with remote SFTP access. Optimized frame pipeline and downscaled video stream to 480p to achieve real-time live object detection at a steady 15 FPS.

Live Performance 15 FPS Real-Time
Edge Model MobileNet-SSD V2
Raspberry Pi Edge AI Computer Vision Python