About Me

I'm a Founding Engineer & AI Architect with 2.5 years of hands-on experience building large-scale AI systems, distributed backend infrastructures, and production-grade full-stack platforms. At ProPeers, I own and engineer the core systems that power 80%+ of total platform traffic including Roadmaps, RoadmapAI, AskAI, CodeLLM, Global AI Search, and the Contextual AI Code Editor.

I design high-scale backend architectures, real-time data pipelines, aggregation engines for 600K+ users, Redis-backed caching layers, search-validation systems, role-based access flows, rate-limiting frameworks, and CI/CD deployment automation that cut release time by 34% and improved reliability across 150+ microservices. Through SSR, dynamic imports and hybrid rendering patterns, I’ve reduced key user journey response times from 1.1s → 200ms, delivering a noticeably smoother product experience.

As an AI Architect, I build Agentic AI pipelines, RAG retrieval systems, MCP protocol layers, and multi-model inference workflows using Azure OpenAI, Azure Databricks, GPT models, and Llama 3.x OSS models. My work spans token optimization, context-window compression, semantic chunking, and adaptive prompt engineering to deliver intelligent experiences at <1s latency under real production traffic.

I’ve engineered RoadmapAI with a self-learning RAG pipeline (text-embedding-ada-002, ChromaDB, semantic filters, vector enrichment), achieving ~99% roadmap accuracy and lifting roadmap ratings from the early 12% baseline. I built CodeLLM, a production AI judge featuring multi-language detection, dual-layer JSON parsing, COMPILATION/RUNTIME/VALIDATION error classification, and deterministic verdict synthesis for educational code evaluation.

I developed AskAI with MCP-layered prompts, resource-type detection (roadmap/article/practice), O1/O3 model routing, token metering, and auto-structured responses improving resolution speed 2× and engagement 3×. I also built the AI Code Editor with ~40ms inference, inline reasoning, multi-language execution, and deep integration with RoadmapAI and CodeLLM, significantly boosting editor retention.

Beyond AI flows, I’ve implemented token-based tiered access systems (one-time/monthly/yearly) on top of these capabilities, and engineered self-optimizing RAG pipelines and distributed multi-model inference workflows that balance accuracy, cost and latency under real-world traffic.

On the product and platform side, I’ve delivered Individual Roadmap Communities, scalable live-stream pipelines, error-resilient API layers, multi-step onboarding flows, connected roadmap progress engines, and search validation systems ensuring hallucination-free retrieval across Roadmaps and RoadmapAI.

At the infrastructure layer, I’ve reduced downtime by 90% (4 hours → 45 mins/month), stabilized Azure VM workloads, eliminated Bastion and high-cost D8 VM footprints, fixed bandwidth cost spikes, and built high-availability fallback layers with cache-first routing and distributed failover.

Day-to-day, I work across MERN + TypeScript, Node.js microservices, Docker/Kubernetes, Azure Cloud, Databricks, CI/CD automation, Prometheus/Grafana observability, and async caching pipelines powering 100K+ monthly active operations.

Outside core engineering, I’m a Problem-Solving & DSA Enthusiast with 5000+ problems solved, a 1700+ day coding streak, and top 0.1% global rankings across platforms. As a mentor to 40,000+ learners, I help engineers master DSA, System Design, Development, DevOps, and Remote Job Preparation, guiding them from theory to real-world success.

I love building scalable systems, intelligent architectures, and next-generation AI-first engineering experiences that blend reliability, performance, and deep technical innovation.