Skip to content
View akop-cyber's full-sized avatar
🎯
Focusing on my class 10th board exams
🎯
Focusing on my class 10th board exams
46 followers · 67 following

Achievements

Achievement: YOLOAchievement: Pull SharkAchievement: Quickdraw

Achievements

Achievement: YOLOAchievement: Pull SharkAchievement: Quickdraw

Highlights

  • Pro

Block or report akop-cyber

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
akop-cyber/README.md

Aarav Kumar Ranjan πŸ˜„

AI/ML Developer and Enthusiast | Independent Researcher | High School Student

I build intelligent systems that bridge the gap between complex algorithms and real-world applications.My core focus is on Reinforcement Learning, intelligent data routing, and scalable backend architecture using Python and FastAPI. I started learning Python in 2024.

πŸ“š Current Research

Systematic Ξ΅-Greedy Bandit

GitHub Β· Zenodo DOI: 10.5281/zenodo.23418529

  • Researched deterministic periodic exploration in stationary Multi-Armed Bandits.
  • Conducted empirical comparisons of deterministic vs. stochastic exploration schedules across 4,000 paired environments.
  • Authored a technical report detailing full methodology, control experiments, and statistical analysis.

πŸ› οΈ Featured Projects

Documind V2

GitHub Β· Live App

  • A lightweight, no-login RAG application for instant PDF Q&A.
  • Implements hybrid retrieval (BM25 + Vector embeddings) combined with Reciprocal Rank Fusion (RRF) for improved context grounding.
  • πŸ“ Note: This is an active learning project optimized for conceptual clarity and demo use, not large-scale production deployment.

Documind

GitHub Β· Live App

  • AI-powered document intelligence platform for processing and analyzing PDFfiles.

RAG Pipeline Evolution (v1 β†’ v3)

GitHub

  • An iterative exploration of RAG architecture, progressing from basic retrieval (v1) to multi-turn conversation history management and advanced LLM integration (v3).

Portfolio AI Backend

GitHub

  • Clean, modular backend architecture built with FastAPI to support my personal portfolio site.

πŸ’» Technical Stack

  • Languages: Python, HTML/CSS
  • Frameworks: FastAPI, PyTorch, Scikit-learn, NumPy, Pandas
  • AI/ML: RAG Systems, LLM Integration, Multi-Armed Bandits, Vector Embeddings
  • Tools: Git, Hugging Face sentence-transformers, Netlify

🎯 Current Focus

  • Exploring Reinforcement Learning
  • Exploring the intersection of Reinforcement Learning and RAG optimization.
  • Mastering advanced retrieval strategies and LLM orchestration.
  • Building robust, well-documented AI backends with FastAPI.

πŸ“ˆ GitHub Stats

GitHub Stats Top Languages

🀝 Connect

I am always open to discussing AI/ML research, collaborating on innovative projects, or connecting with fellow developers.

Last Updated: October 2026

Pinned Loading

  1. systematic-epsilon-greedy-bandit systematic-epsilon-greedy-bandit Public

    An empirical comparison of deterministic vs. stochastic exploration in stationary Multi-Armed Bandits.

    Python 1

  2. Backpropagation_model Backpropagation_model Public

    This code is written purely in python to demonstrate the working of backpropagation using a 2-2-2 Neural Network

    Python 1

  3. Documind-V2 Documind-V2 Public

    Chat Instantly to your pdfs. Just PLUG AND PLAY no unncessary drama

    Python 2

  4. RAG-built-from-scratch RAG-built-from-scratch Public

    this is an rag which is built from scratch

    Python 3

  5. Linear-Regression-from-Scratch Linear-Regression-from-Scratch Public

    Self coded linear regression model

    Python 2

  6. Logistic-Regression-Model-from-Scratch Logistic-Regression-Model-from-Scratch Public

    This Logistic Regression model is coded from scratch and trained on Inosphere dataset

    Python 1