B.Tech CSE · KIIT · 2027

Aashish
Sagar

I build backend systems that don't leak your data.

From FHE-encrypted AML pipelines to AI-powered dev tools — currently grinding Striver's DSA sheet with 150+ problems solved. Endgame: Google SWE.

KIIT, Bhubaneswar
CGPA 8.86
LeetCode 1440
aashish@dev — zsh
aashish@dev:~$ cat profile.json
{
  "name": "Aashish Sagar",
  "role": "Full Stack Developer",
  "cgpa": 8.86,
  "leetcode": 1440,
  "problems_solved": "150+",
  "goal": "Google SWE",
  "status": open_to_opportunities,
  "email": sagaraashish25@gmail.com
}
aashish@dev:~$

Who I Am

Hey! I'm a B.Tech CSE student at KIIT, Bhubaneswar (2023–2027), currently in my 3rd year, maintaining a CGPA of 8.86. My passion is building backend-heavy systems that solve real, complex problems.

I started with low-level C and data structures — sparse matrices, linked lists, recursion — and grew into shipping production-grade systems: Spring Boot microservices, FHE-encrypted pipelines, and agentic LLM workflows.

I've contributed to open source via GSSoC and SSoC, merged PRs into active production codebases, and co-authored technical documentation that onboards hundreds of contributors.

Currently deep in Striver's DSA sheet with 150+ problems solved and a LeetCode rating of 1440. The endgame is Google SWE — every line of code gets me closer.

0 CGPA (/ 10)
0 DSA Problems
0 Open-Source PRs
0 Real-World Systems
KIIT University
2023 – 2027
B.Tech in Computer Science & Engineering
CGPA: 8.86 / 10  ·  Bhubaneswar, Odisha

Experience

Apr 2025 – Jun 2025
  • Triaged and resolved backend bugs across 2+ active repositories, delivering targeted fixes via pull requests merged into production codebases maintained by industry contributors.
  • Drove 4+ issues from root cause analysis all the way to merged PR — covering bug fixes, feature additions, and test coverage improvements.
  • Restructured technical documentation to accelerate contributor onboarding, standardizing Gitflow-based version control workflows used across the project.

Tech Skills

// Languages
Java TypeScript JavaScript Rust C SQL Python
Java90%
TypeScript / JS78%
C / Systems82%
// Backend & Infra
Spring Boot Spring MVC WebFlux LangChain4j REST APIs MySQL Docker Railway
Spring Boot84%
Microservices & APIs80%
// Frontend
React.js HTML5 CSS3 Java Swing
HTML / CSS92%
React.js72%
// CS Fundamentals
DSA (Java) OOP System Design Multithreading Git / GitHub LeetCode
DSA (Java)80%
OOP / System Design86%
// Security & Crypto
FHE (TFHE-rs) JNI (Rust↔Java) HMAC-SHA256 Rate Limiting Privacy-Preserving Comp.
FHE / Cryptography70%
Security Patterns78%
// AI / ML
LangChain4j Groq LLaMA Agentic Pipelines CNN / OpenCV Flask
LLM Integration75%
ML / Computer Vision65%

What I've Built

Project 001 · Flagship
CIPHER — Agentic-FHE AML Gateway
A privacy-preserving Anti-Money Laundering system where transaction data is never decrypted at any stage. Built as a 3-tier microservices pipeline combining Fully Homomorphic Encryption with an autonomous LLM compliance agent.
FHE Engine: Custom Rust-to-Java JNI bridge using Zama TFHE-rs classifies encrypted transactions into 4 risk tiers in 50–55ms with zero plaintext exposure.
Agentic AI: LangChain4j + Groq Llama 70B autonomously routes flagged transactions — reducing verdict latency to 3–8s end-to-end.
Reliability: 100% pass rate across 14 end-to-end test cases.
Java Spring Boot Rust + JNI Groq LLaMA 70B React TypeScript
Project 002 · Deployed
CodeReview Bot — AI-Powered PR Reviews
A stateless, event-driven service that plugs into GitHub webhooks and automatically reviews pull requests using AI. The full pipeline — from diff fetch to AI analysis to posted comment — completes in under 4 seconds.
Event-Driven: Webhook-driven Spring Boot service intercepts PR events, fetches diffs via GitHub REST API, and posts structured feedback automatically.
4-Layer Security: HMAC-SHA256 signature verification, per-repo token-bucket rate limiting (Bucket4j), prompt injection guards, and a global exception handler.
Production-Ready: Multi-stage Docker build deployed on Railway with health-check-based auto-restart.
Java Spring Boot Groq LLaMA 3.3 Docker WebFlux
Project 003
Real-Time Image Classifier
A live webcam-based visual classification system built around a trained CNN model. Frames are streamed via OpenCV to a Flask backend that runs inference and returns predictions in real time.
CNN Pipeline: Trained and deployed a convolutional neural network for multi-class image recognition.
Real-Time Stream: OpenCV handles frame capture and preprocessing at the hardware level, piped to Flask for low-latency inference.
Latency: Achieves near real-time inference with optimized preprocessing pipeline.
Integration: Seamlessly connects OpenCV capture with Flask API for continuous prediction flow.
Python CNN / ML Flask OpenCV

Achievements

LeetCode Rating 1440
Solved 150+ DSA problems across competitive platforms, working through Striver's A2Z DSA sheet systematically.
FHE in Production
One of the few developers to ship Fully Homomorphic Encryption in a working product — encrypted financial data end-to-end, zero plaintext exposure.
GSSoC & SSoC Contributor
Merged 4+ pull requests into active open-source production repos during two concurrent programs in 2025.
CGPA 8.86 / 10
Maintaining strong academics at KIIT alongside building production-grade projects and open-source contributions.
Agentic AI Systems
Built a fully autonomous compliance pipeline using LangChain4j + Groq Llama 70B with sub-8s automated verdicts.
Deployed to Production
CodeReview Bot is live on Railway with multi-stage Docker containerization, HMAC security, and health-check auto-restart.

Let's Connect

Whether it's an internship opportunity, a hackathon team, or just a deep dive into distributed systems and cryptography — my inbox is always open. I respond within 24 hours.

// Current Status
Available for Opportunities
✓  Summer internships (2026)
✓  Open-source collaborations
✓  Hackathon team-ups
✓  Research partnerships
✓  Freelance backend projects

Response time: < 24 hours
Phone: +91 9234204036