██╗ ██╗███████╗███████╗██████╗ ████████╗██╗ ██╗██╗ ██╗ ██╗ █████╗ ███████╗ █████╗ ███╗ ██╗
██║ ██╔╝██╔════╝██╔════╝██╔══██╗╚══██╔══╝██║ ██║██║ ██║ ██║██╔══██╗██╔════╝██╔══██╗████╗ ██║
█████╔╝ █████╗ █████╗ ██████╔╝ ██║ ███████║██║ ██║ ██║███████║███████╗███████║██╔██╗ ██║
██╔═██╗ ██╔══╝ ██╔══╝ ██╔══██╗ ██║ ██╔══██║██║ ╚██╗ ██╔╝██╔══██║╚════██║██╔══██║██║╚██╗██║
██║ ██╗███████╗███████╗██║ ██║ ██║ ██║ ██║██║ ╚████╔╝ ██║ ██║███████║██║ ██║██║ ╚████║
╚═╝ ╚═╝╚══════╝╚══════╝╚═╝ ╚═╝ ╚═╝ ╚═╝ ╚═╝╚═╝ ╚═══╝ ╚═╝ ╚═╝╚══════╝╚═╝ ╚═╝╚═╝ ╚═══╝
name: Keerthi Vasan
role: Backend Architect · Enterprise AI · Cloud-Native SDE
education: CCE @ Amrita Vishwa Vidyapeetham
location: Tamil Nadu, India
engineering_focus:
- High-Throughput Microservices (Spring Boot 3, Go, Kafka)
- Agentic AI Systems & Smart RAG (Spring AI, Bedrock, PGVector)
- Cloud Infrastructure & Kubernetes (GKE, AWS EC2/Lambda, Docker)
problem_solving:
- System-First Problem Solving (Distributed Trade-offs, CAP & Resilience)
- Advanced Data Structures & Algorithms (150+ Solved on LeetCode @Kvasan)
- Competitive Programming & Algorithmic Reasoning
open_to:
- High-Impact Backend & SDE Engineering Roles
- Enterprise AI & Cloud Native Systems Architecture
- Solving Complex, Scalable Real-World Engineering Problems
I sit at the intersection of scalable backend systems and intelligent AI workflows.
✦ System-First Mindset: I approach software engineering from a first-principles architectural perspective — evaluating distributed trade-offs (CAP theorem, latency vs. consistency), data partitioning, and fault-tolerant boundaries before writing code.
✦ Engineering Philosophy: Deconstruct complex real-world problems into modular, high-throughput microservices that are resilient, scalable, and self-improving.
|
|
|
|
|
|
|
|
|
|
|
Voice-first agentic civic automation for rural India via Amazon Connect IVR & Web Phone Emulator. Uses Amazon Kendra Smart RAG for document retrieval, AWS Lambda webhooks, AWS Bedrock (Claude 3.5), and a deterministic rules engine. Stack: AWS Bedrock · Kendra (RAG) · AWS Lambda · Connect · Next.js 16 · Python
|
MCP-powered agentic AI platform for real-time manufacturing root-cause analysis. Orchestrates 4 specialized AI agents with Model Context Protocol (MCP) and Nitrostack to convert fragmented MES/sensor data into evidence-backed corrective actions. Stack: Model Context Protocol (MCP) · Nitrostack · Gemini · Python · FastAPI · React
|
|
Production biometric face attendance & real-time campus geofencing platform for Amrita Vishwa Vidyapeetham. Features ArcFace 512-D ONNX embeddings, PGVector similarity search, MiniFASNetV2 anti-spoofing, and Ray-Casting PiP geofencing. Stack: Python · FastAPI · PyTorch/ONNX · PGVector · Next.js 15 · Leaflet.js
|
Enterprise High-Throughput Concurrent File Distribution Engine & Zero-VPN Campus Intranet Mesh Relay Gateway. Features parallel 1-to-10 FTP/FTPS fan-out delivery with cryptographic SHA-256 integrity, WebSocket reverse tunneling for remote firewall bypass, and an on-demand AWS EC2 cost controller. Stack: Java 21 · Spring Boot 3 · MongoDB Atlas · WebSocket Reverse Tunneling · Docker · AWS EC2
|
|
Enterprise-grade multi-channel notification engine powering Klarixa AI's GxP compliance SaaS platform. Features 4-tier Redis priority queues, sliding-window rate limiting, circuit breaker fault tolerance, and Jinja2 templating. Stack: FastAPI · Python 3.11 · PostgreSQL · Redis · Docker · AWS EC2
|
Production-grade Spring Boot microservices integrated with AWS Bedrock & PGVector. Orchestrates LLMs (GPT-4o, Claude, Llama), high-throughput RAG search, and real-time Kafka event streaming. Stack: Java 21 · Spring Boot 3 · Spring AI · AWS Bedrock · PGVector · Kafka
|
| Year | Role / Milestone | Domain & Technical Impact |
|---|---|---|
2026 |
✦ AI Engineer Intern @ Sparks AI | Multi-tenant CommandOS platform, PostgreSQL RLS, OAuth2 PKCE & 429 Vitest tests |
2026 |
✦ DevOps Lead @ NIT Trichy | RHEL 9 server infrastructure, systemd services & deployment |
2026 |
✦ Google Certified GKE Architect | Enterprise Kubernetes workloads, Helm & cluster architecture |
2025 |
✦ AI Engineer Intern @ Infosys | Production GenAI systems, RAG search & LLM orchestration |
2025 |
✦ Built LeetCode Progress Tracker | Next.js 15 + Supabase full-stack interview prep engine |
2025 |
✦ 150+ DSA Problems Solved | Active competitive problem solver on LeetCode (@Kvasan) |
2024 |
✦ IoT + AI Integrated Systems | Applied computer vision, sensor telemetry & edge AI |
2024 |
✦ Codefest Competitor | Competitive programming circuit & algorithm design |
🚀 My Custom Engineering Project: Engineered LeetCode Progress Tracker — a full-stack interview prep engine that automates problem metadata fetching via GraphQL, manages company-wise DSA sheets (Google, Meta, Amazon), and persists user progress with Next.js 14 & Supabase.
