About Me
I'm a builder who takes products from early uncertainty to production-grade systems that scale.
I was the founding CTO of Plutus (raised from A16Z & Konvoy Ventures), where I built a Real Money Gaming & Prediction markets platform from scratch, scaling to 16k users within 3 months. I managed our entire AWS infrastructure and navigated rapid platform migrations under tight constraints.
I also led the 0→1 development of Pixelsurf, an AI-native game creation platform, taking it from idea to MVP and enabling users to generate, publish, and monetize online games using AI.
Previously, I was a 2x Google Summer of Code contributor for OpenCV, implementing speech recognition via AudioIO. My technical depth includes building ML/DL models from scratch (Deep Learning, GANs, RL, Computer Vision), implementing automated Quantization Aware Training for IoT devices, and designing real-time gesture recognition systems at Siemens.
I'm an IIT Kanpur graduate where I self-taught CS and AI while majoring in Civil Engineering—a path that taught me how to learn hard things independently and think from first principles.
Journey
Co-Founder & CTO at Plutus Labs
Raised from A16Z & Konvoy Ventures
Built a Real Money Gaming & Prediction markets platform from scratch, scaling to 16k users within 3 months while managing multiple platform pivots & fast paced feature development.
- →Scaled to 16k users in 3 months post-launch
- →Designed & Managed the entire AWS infrastructure & automated CI/CD pipeline
- →Led platform migrations: Mobile → Web → Mobile, achieved a p95 latency of <1s for all APIs
- →Reduced time-to-interactive from 10-12s to <2s while supporting 5x user growth
- →Built real-time gaming infrastructure with WebSocket orchestration in a distributed system
- →Led the development of payment systems, anti-fraud mechanisms, analytics and auto banning systems
- →Designed and built the MVP of pixelsurf, an AI-native game creation platform
- →Managed a team of 6 engineers
AI Software Engineer at RISA Labs Inc.
Worked on early-stage AI systems in healthcare, spanning insurer-facing analytics and a proof-of-concept for automating oncology pre-authorization using multi-agent LLM pipelines and unstructured medical data.
- →Built a proof-of-concept population health analytics dashboard to support insurance partner onboarding and patient-level insights
- →Designed a multi-agent AI pipeline to automate oncology pre-authorization based on insurance coverage and clinical guidelines
- →Implemented RAG-based ingestion of highly unstructured medical data (labs, physician notes, reports)
- →Delivered a PoC targeting reduction of first-pass pre-authorization denials from >50% to <30%
Co-Founder at Unicohub / Plutus
Co-founded and built a creator fan platform from idea to MVP, enabling users to interact with their fans through live streams and chat.
- →Led 0→1 product development from scratch
- →Built full-stack platform with modern web technologies
- →Integrated multiple LLMs and AI models
- →Later pivoted to Plutus Labs
Contributor, OpenCV
Google Summer of Code 2022
Contributed to OpenCV as GSoC participant, implementing speech recognition capabilities via AudioIO module.
- →Created a Audio Visual Speech Recognition Sample in OpenCV, ensembling two models trained on different modalities.
- →Studied various Lip reading models and exported the pytorch TCN model into ONNX format by editing the model graph.
- →Written and Merged production level code to inference ONNX model using DNN module in official OpenCV repo.
Research Intern at Siemens Technology India
Designed and implemented a real-time dynamic hand gesture recognition model for industrial IoT applications.
- →Constructed NVIDIA’s Tao compatible dataset to train gesturenet classifying static hand gestures with ~80% accuracy
- →Trained DeepGRU and TCN models on DHG dataset with average top 1 accuracy of 90% on validation split each
- →Augmented hand skeleton data to increase data amount for training in a domain and view point agnostic setting
- →Recorded & Augmented data with mediapipe to train TCN model for dynamic hand gesture classification in industries
AI Intern at EDGENeural.ai
Developed edge AI solutions for resource-constrained devices and IoT applications.
- →Automated training for object classification models using MMClassification framework in a docker container
- →Implemented Quantize Aware Training in torch for models such as efficientNet, mobileNet, shufflenet etc
- →Created custom callback functions to save and send training logs and dataset info. to frontend via REST API
Student Developer, OpenCV
Google Summer of Code 2021
Created speech recognition sample in OpenCV using AudioIO module, implementing multimodal data processing.
- →Created a Speech Recognition Sample in OpenCV, making it the very first speech processing application in OpenCV.
- →Studied various State of The Art Automatic Speech recognition models and used NVIDIA’s Jasper (arXiv:1904.03288v3) model to create pre-trained onnx file compatible with OpenCV.
- →Created pre-processing functions to extract features & decoding functions to produce transcript using Numpy
- →Edited the computation graph of Onnx file to fix dependency issues in OpenCV.
Summer SDE Intern at Monetix Ltd.
Worked on full-stack development projects, extending backend systems and building mobile features.
- →Extended existing backend written in Typescript & frontend in Ionic & Angular to streamline new partner onboarding.
- →Created a conditional beta build system & added new partners’ support in Ionic & Angular app
- →Automated notification & mail delivery for payment reminders in vanilla JS on Heroku backend using Node CRON.
- →Improved Angular Frontend to provide credit history for loan approval.
- →Scripted data migration to & from MongoDB to take backup & restore data in production equipped with data dumping.
Bachelor of Technology, Civil Engineering
Indian Institute of Technology, Kanpur
Self-taught Computer Science and AI while majoring in Civil Engineering. Built strong foundation in ML/DL, implementing models from scratch.
- →Self-taught CS fundamentals, algorithms, and system design
- →Implemented ML/DL models from scratch (CNNs, GANs, RL, Computer Vision)
- →Built and trained custom neural architectures
- →Focused on deep learning theory and practical implementation
- →Active in coding competitions and technical projects