Ruthuraraj ML & AI Lab
Change the repository type filter
All
Repositories list
27 repositories
.github
Public- An AutoGen-powered Financial Intelligence Platform that transforms bill images into actionable insights through OCR, expense analytics, forecasting, and multi-a…
- Memory-driven multi-agent research system built with LangGraph, RAG, web search, reflection loops, evidence curation, and persistent memory for explainable rese…
- Build, analyze, and visualize a custom Byte Pair Encoding (BPE) tokenizer trained on the WikiText-2 dataset with vocabulary analysis, compression evaluation, an…
- Implemented Parameter-Efficient Fine-Tuning (PEFT) using LoRA and QLoRA on BERT and Gemma 2B, achieving memory-efficient sentiment classification while training…
- End-to-end sentiment analysis project comparing RNN, LSTM, GRU, and BERT on over 1 million Twitter posts using PyTorch and Hugging Face Transformers.
- An industrial-grade multi-agent decision support system (DSS) built with CrewAI Flows and Gradio. Automates real-time inventory risk vectors, transit route opti…
- An advanced multi-agent orchestrator utilizing LangGraph StateGraph, conditional reflection routing, and ChromaDB RAG memory to generate self-correcting market …
github.io
Public- A cognitive agent architecture using LangGraph and Python custom orchestration for adaptive travel planning. Employs a non-destructive state machine with dynami…
ReAct-Web-Research-Agent
PublicAn autonomous AI research assistant built from scratch using the ReAct (Reasoning + Acting) paradigm. Features a cost-aware multi-LLM routing engine (Gemini 3.1…- A heuristic-based ReAct-lite RAG agent that answers questions grounded in workshop materials — built as a live demo for a 3-day AI workshop at SNS College of Te…
- AI Content Studio: A multi-agent AI system where research, illustration, and reviewer agents collaborate to generate educational explanations and diagrams.
Embedding_Search
PublicAn end-to-end Word2Vec architecture (Skip-Gram with Negative Sampling) built entirely from scratch in PyTorch and deployed as an interactive, browser-based visu…- A implementation of generative computer vision models reconstructing high-resolution images from low-resolution inputs using SRGAN and ESRGAN (RRDB-based backbo…
- An architectural exploration tracking the progression of automatic caption generation, moving from traditional deep learning frameworks to modern multimodal arc…
ProTip! When viewing an organization's repositories, you can use the
props. filter to filter by custom property.