I am a final year B.Tech Computer Science and Engineering student at IIIT Bhagalpur, graduating in 2027. I enjoy solving challenging Data Structures & Algorithms problems and building scalable full-stack and AI-powered applications.
My interests span Full-Stack Development, Generative AI, RAG, Multi-Agent Systems, Microservices, System Design, and Event-Driven Architecture.
#include <iostream>
#include <vector>
struct Developer {
string name = "Ravi Prakash Pal";
string education = "B.Tech CSE @ IIIT Bhagalpur (2027)";
string cgpa = "8.05 / 10";
string recent_role =
"Software Engineer Intern @ Knockhere Technologies";
vector<string> interests = {
"Competitive Programming",
"Full-Stack Development",
"Generative AI & RAG",
"System Design & Microservices",
"Event-Driven Architecture"
};
void currentStatus() {
cout << "Building scalable full-stack & AI-powered systems!" << endl;
cout << "Exploring RAG, Multi-Agent Systems & Microservices." << endl;
cout << "Actively targeting Software Engineering Internships!" << endl;
}
};
B.Tech in Computer Science and Engineering π September 2023 β June 2027 π CGPA: 8.05 / 10
May 2025 β September 2025
- Developed a production-grade food delivery PWA using React, Supabase, and Node.js, featuring 15+ pages, 35+ reusable components, OTP authentication, and real-time order tracking.
- Built a secure Razorpay payment pipeline with JWT verification and SHA-256 tamper detection, supporting wallet, COD, and online payments across 8 order lifecycle states.
- Optimized the data access layer by authoring 15+ PostgreSQL RPC functions and implementing service-worker caching across 20+ tables, enabling offline-first functionality and reducing redundant network requests.
AI-powered travel planning assistant using LangChain and LangGraph
- Architected an LLM-powered smart chatbot in Python using LangChain and LangGraph, orchestrating a Supervisor agent to dynamically route requests to specialized AI sub-agents.
- Built a RAG (Retrieval-Augmented Generation) pipeline with ChromaDB, using text embeddings and NLP techniques to improve contextual retrieval and reduce hallucinations.
- Engineered strict input guardrails and a Human-in-the-Loop (HITL) workflow for prompt safety and controlled AI interactions.
- Integrated PostgreSQL for structured data management and maintained the project using Git/GitHub.
Tech Stack: Python LangChain LangGraph ChromaDB PostgreSQL RAG Guardrails HITL
Microservices-oriented EdTech platform for course sales and video streaming
- Delivered a full-stack EdTech platform featuring 13 pages, 9 data models, and 37+ REST APIs for course sales and video streaming.
- Integrated Razorpay for payments, Cloudinary for media management, and implemented 3 role-based access levels.
- Migrated OTP storage from MongoDB to Redis, reducing latency by approximately 95% (~1ms); introduced Redis-backed JWT blacklisting and multi-layer rate limiting.
- Refactored the monolith into User and Mail microservices using RabbitMQ, reducing OTP API latency by 40% through asynchronous SMTP offloading.
Tech Stack: JavaScript React.js Node.js Express.js MongoDB Redis RabbitMQ Razorpay Cloudinary
Dual-role job marketplace connecting candidates and employers
- Engineered a dual-role job board using Next.js 16, TypeScript, Prisma, and NeonDB, supporting 10+ data models, Stripe subscriptions, and NextAuth v5 with GitHub/Google OAuth.
- Integrated GPT-4o-mini for automated resume evaluation across 5 weighted criteria, powering a 5-stage hiring pipeline.
- Implemented event-driven email notifications using Inngest and Resend.
- Strengthened application security with Arcjet bot detection and rate limiting across 12 server actions, Stripe webhook signature validation, and Zod schema validation across 6 API input schemas.
Tech Stack: Next.js 16 TypeScript Prisma NeonDB Stripe NextAuth v5 Inngest Resend Arcjet OpenAI API Zod
- π LeetCode Knight: Peak rating 1851, peak global rank 1779.
- π‘ CodeChef 3-Star: Peak rating 1606, peak global rank 437.
- π΅ Codeforces Pupil: Peak rating 1291, peak global rank 1459.
- π§ 1200+ DSA / Competitive Programming problems solved.
- π Bit-by-Bit Hackathon: Secured 7th place in a 36-hour hackathon as part of a 5-member team.
- βοΈ Competed in 30+ rated contests across competitive programming platforms.
| π₯ Area | π― Focus |
|---|---|
| DSA & Competitive Programming | Solving Medium/Hard problems and maintaining competitive programming consistency |
| Generative AI | Building RAG pipelines, Multi-Agent Systems, Guardrails & HITL workflows |
| System Design | Designing Microservices, asynchronous systems, caching & event-driven architectures |
| Full-Stack Development | Building production-ready applications with Next.js, React, Node.js & PostgreSQL |
| Backend Engineering | Exploring Redis, RabbitMQ, REST APIs and scalable service architectures |


