CS + AI & Data Science · IIIT Kottayam '27

Building AI
that ships.

I'm Akshay — I build end-to-end systems across computer vision, full-stack engineering, and applied ML. From hackathon floors to research labs.

Akshay Sriram
Bhaskara Akshay Sriram
CS + AI/DS · Class of '27
📍 Hyderabad, Telengana, India
🎓 IIIT Kottayam · B.Tech CSE + AI/DS
Azure AI Certified
scroll
PythonTypeScriptReactFastAPIYOLOv8OpenCVNext.jsPostgreSQLTensorFlowQiskitAzure AIC++ PythonTypeScriptReactFastAPIYOLOv8OpenCVNext.jsPostgreSQLTensorFlowQiskitAzure AIC++

Things I've built.

Varroc Eureka 3.0 Hackathon · 2026
01
Two-Wheeler Driving
Behavior Scorer

Real-time ADAS system scoring riding behavior 0–100 with live video analysis. Computer vision pipeline with 60fps WebSocket streaming, Kalman filter on Lucas-Kanade optical flow, calibration-free speed estimation, and 6 behavioral metrics including helmet detection via Otsu blob analysis.

PythonYOLOv8OpenCVFastAPIReactWebSocketKalman Filter
View on GitHub Read the deep-dive
02
Cal Clone

Production-grade Cal.com clone. Normalized 6-table PostgreSQL schema, real-time double-booking prevention, automated email reminders via Cron, 11 REST endpoints. Deployed on Vercel.

Next.js 14TypeScriptPostgreSQLPrismaTailwind
View on GitHub
03
Quantum Medical Diagnostics

Hybrid quantum-classical deep learning model combining CNNs with variational quantum circuits using 4-qubit angle encoding. 6–8% accuracy improvement → 90% classification accuracy.

PythonTensorFlowQiskitKeras
View on GitHub
04
Non-Invasive Diabetes Detection

Research under Dr. Mahesh Parihar, IIT Bombay. Classification models on the Pima Indians Diabetes Dataset achieving 96% accuracy with full ROC-AUC benchmarking.

PythonScikit-learnML
View on GitHub
05
Airplane Engine Failure Estimation

Student researcher under Dr. Yogeshwar Singh, DIAT Pune. ML models for Remaining Useful Life (RUL) estimation on the C-MAPSS dataset with a full preprocessing dashboard.

PythonMLC-MAPSS
View on GitHub

What I work with.

Languages
C++PythonTypeScriptJavaScriptJavaSQL
AI / ML
YOLOv8OpenCVTensorFlowQiskitScikit-learn
Web
ReactNext.jsFastAPINode.jsTailwindPrisma
Cloud & Tools
Azure ✓PostgreSQLGitDockerPostman
Core CS
OOPSOSDBMSSystem DesignNetworks
Competitive
350+ LeetCodeDSAC++

Builder by nature.

I'm a CS + AI/DS undergrad at IIIT Kottayam (Class of 2027) who builds end-to-end AI systems — from computer vision pipelines and real-time backends to full-stack dashboards.

My work spans automotive, healthcare, aerospace, and education. I care about shipping solutions that are technically rigorous and practically useful — not just demos.

I actively compete in hackathons and programming contests, and I've conducted research under faculty at IIT Bombay and DIAT Pune.

When I'm not shipping code I'm under the barbell — won the IIIT-K Powerlifting meet in 2026.

350+
LeetCode problems
5+
Projects shipped
2
Research projects
2027
Graduation year
Microsoft Azure AI Engineer Associate
Certified · Valid July 2025 – July 2026

Technical deep-dives.

I write about what I build — the architecture decisions, the bugs that took hours to fix, and the insights that only come from shipping real systems. Published on Medium.

Computer Vision
ADAS
2026 · 8 min read · Medium
Building a Real-Time Two-Wheeler Driving Behavior Scorer: From Hackathon Idea to Production ADAS System
How I went from a blank Python file to a full-stack ADAS system in a hackathon window — YOLOv8 object detection, a Constant-Acceleration Kalman filter, Lucas-Kanade optical flow for ego-motion compensation, calibration-free MPP speed estimation, and 6 real-time behavioral scoring metrics. Shipped with a React + FastAPI dashboard streaming at 60fps over WebSocket.
YOLOv8 + OpenCV
Kalman Filtering
FastAPI + WebSocket
React Dashboard
Read on Medium
More writing on Medium →

Let's build
something real.

Got a project, opportunity, or just want to talk tech?

akshaysriram.b@gmail.com