FULL-STACK DEVELOPER | AI ENGINEER | DSA ENTHUSIAST

Alok Kumar
builds scalable web applications
end to end.

I build full-stack applications and AI-powered solutions, combining practical software engineering with RAG, LLMs, and intelligent automation.

โ— Available for software engineering and AI/ML / GenAI internships

8.3CGPA / 10
650+LeetCode Solved
1606CodeChef Highest
Open SourceContributor
Alok Kumar
๐Ÿ“ Noida, India
01 / ABOUT

What I'm working with

B.Tech CSE student graduating in 2028, with hands-on experience in full-stack development, GenAI, and teaching DSA.

EDUCATION

B.Tech, Computer Science and Engineering

ABES Engineering College, AKTU, Ghaziabad
2024 โ€“ 2028 ยท Current CGPA 8.58/10

DSAOOPSoftware DesignOperating Systems

Bihar School Examination Board, Patna: Senior Secondary (PCM) 88% ยท High School 90%

FOCUS
AI + Full-Stack

RAG, LLMs and agentic workflows, built on a MERN foundation

PROBLEM SOLVING

650+ on LeetCode

CodeChef highest rating: 1606

WHAT I BUILD

Full products, not just models

I build with React, Node.js, Express, REST APIs and databases, and add AI features like RAG, OCR, embeddings, vector databases and LLM integration.

TEACHING

Teaching Assistant, DSA (Java)

Apna College, Feb โ€“ Jun 2026. Mentoring students through live coding, dry runs and doubt solving.

02 / SKILLS

Toolbox

Full-stack engineering to build the product, AI to make it smart.

PROGRAMMING
JavaPythonC++JavaScriptSQL
CS FUNDAMENTALS
DSAOOPDBMSOperating SystemsComputer NetworksSoftware Design
FRONTEND
React.jsHTMLCSSViteTailwind CSS
BACKEND
Node.jsExpress.jsFastAPIFlaskREST APIsMVC Architecture
DATABASES
MongoDBMySQLPostgreSQLSupabase
TOOLS & DEPLOYMENT
GitGitHubDockerPostmanAWS (EC2, S3, IAM)NetlifyRenderCI/CD
AI / GENERATIVE AI
RAGOCREmbeddingsLangChainLangGraphQdrantChromaGroq APIOpenCVMediaPipeLLMsAgentic AIMachine LearningDeep LearningNLPTransformersReinforcement LearningPyTorchScikit-learnHugging Face
03 / PROJECTS

Selected work

RAG, computer vision, agentic automation, reinforcement learning, and full-stack MERN applications.

IKIP: RAG Knowledge Platform

A RAG-based document question-answering platform: upload documents, then ask questions and get answers grounded in their content.

  • OCR extracts text from documents, which is embedded with FastEmbed
  • Embeddings are stored in Qdrant for vector search and retrieval
  • LangChain and a Groq LLM answer questions from the retrieved context
RAGOCRFastEmbedQdrantLangChainGroq

AI Real-Time Gym Coach

May โ€“ Jun 2026

Real-time form analyser tracking 33 body joints to catch bad posture as you lift.

  • Tracks 33 joints with MediaPipe + OpenCV, posture feedback in under 100ms
  • Auto rep counter, gTTS voice feedback, Groq API tips. Only correct-form reps count
  • Workout history stored in Supabase with login; deployed live at gymai.aloktech.in
PythonMediaPipeOpenCVGroq APISupabase

RAG Document Q&A System

Apr โ€“ May 2026

Retrieval pipeline that answers questions across a 50+ document library.

  • Chunked documents into 1,000+ embeddings stored in ChromaDB for fast search
  • Groq API returns answers in under 2 seconds across 20+ document types
  • Built with LangChain, Sentence Transformers, and NLP-based retrieval
LangChainChromaDBGroq APISentence Transformers

AI Agent for Email Automation

Mar โ€“ Apr 2026

Agentic workflow that reads an instruction, drafts an email with an LLM, and sends it.

  • Step-by-step agent: understands the request, decides the action, executes it
  • Automated 50+ email tasks, cutting manual writing time from 5 min to under 15 sec
Agentic AILLMs

Deep RL: Flappy Bird (DQN)

Jan โ€“ Feb 2026

A DQN agent trained from scratch to master Flappy Bird through trial and error.

  • Trained over 50,000+ episodes with reward shaping and exploration strategies
  • Improved from scoring 0โ€“2 pipes randomly to clearing 10+ pipes after training
Reinforcement LearningDQNDeep Learning

StayGlory: Hotel Booking Platform

Aug โ€“ Nov 2025

A full-stack MERN hotel booking app, built and deployed end to end.

  • Built with 10โ€“30+ hotel listings using the MERN stack, following MVC architecture for REST APIs
  • Secure login and session handling across pages with Passport.js + MongoDB
  • Mapbox for location search, Cloudinary for image upload and storage
MERNPassport.jsMapboxCloudinaryREST APIs

GlowGirl: Women's Product Shopping Platform

Jan โ€“ Feb 2026

A full-stack MERN e-commerce platform for women's products with order tracking.

  • Lists 40โ€“50 products with a full admin panel to manage products, orders, and users
  • Improved app performance by 20% by optimizing MongoDB queries for product listing
  • Cart, checkout, and Cash on Delivery flow, with 15+ successful order deliveries on the live site
MERNMongoDBAdmin PanelREST APIs
04 / EXPERIENCE

Work & impact

February 2026 โ€“ June 2026

Teaching Assistant: Data Structures & Algorithms (Java)

Apna College ยท Remote
  • Mentored 5,000+ students in Java and DSA through live coding and interview-focused sessions
  • Resolved 162+ technical doubts on Java and DSA concepts, achieving an average student rating of 4.12/5
  • Conducted debugging sessions and algorithm dry runs to strengthen problem-solving skills
  • Prepared technical explanations and learning resources with the course instructor
05 / CODING PROFILES

Problem solving

LEETCODE
650+

problems solved โ†—

CODECHEF
1606

highest rating

06 / ACHIEVEMENTS

Certifications & recognition