My GitHub profile presents me as a machine learning enthusiast progressing toward an ML engineering role, focused on deploying real-world ML systems. I joined GitHub in 2024. The profile lists 1 follower and shows Python (9) and Jupyter Notebook (6) across my repositories.
My GitHub README presents me as a B.Tech Computer Science — Data Science student in India, focused on machine learning, generative AI, and practical software. I describe my strength as “Bridging the gap between raw data and intelligent decisions” and my approach as “learning by shipping”: I build projects to connect concepts with working real-world applications. My profile motto is “Engineer in progress. Always building. Always learning.”
The README lists my interests as machine learning and model building, generative AI and large language models, RAG systems, data analysis and visualization, and scalable AI APIs with FastAPI. My stated current activities are building an AI-powered data analyst web app, deepening my LangChain and vector-search knowledge, and learning MLOps and cloud-deployment patterns. I describe my work as spanning end-to-end ML pipelines, RAG-based document intelligence, and production-ready APIs built with maintainable Python; the README does not give implementation details or results for those efforts.
My README has Tech Stack sections for Languages, AI / ML / Data, Backend & Databases, and Tools & DevOps, but lists no technologies beneath them. Its GitHub statistics widgets are configured for Aakashkumar017; the stats request all commits, private contributions, icons, and a GitHub rank icon, while the compact top-languages widget requests private-repository counts and 8 languages. The stats images use widths of 49%, 49%, and 40%; the header image is 180 pixels high with font size 42 and description size 16, and the footer is 100 pixels high. The widgets use the tokyonight theme and hide borders. The Contribution Graph section is empty, so no contribution data is shown. The README closes with “Without data you're just another person with an opinion.”, attributed to W. Edwards Deming.
My applied ML projects include ML-Diabetes-Prediction-API, a production-ready FastAPI and Docker backend that uses AdaBoost to predict diabetes risk from medical parameters; a food-delivery-time-prediction application using rider, order, traffic, weather, vehicle, distance, and timing features; and a loan-prediction repository. The food-delivery project is also documented in deliveryeta food delivery time prediction. The diabetes repository has 1 star.
My AI and retrieval projects include RAG-PDF-Chatbot, which combines LangChain, FAISS, Hugging Face Embeddings, and Groq; Quantum-RAG-Evaluation, which adds automated evaluation of semantic similarity, keyword coverage, BLEU score, and retrieval quality; and a PDF claim-verification platform using FastAPI, Streamlit, Groq Llama 3.3, Tavily Search, and PyMuPDF for claim extraction, verification, confidence scoring, and evidence analysis. I also list a conversational AI system built with FastAPI microservices, a PyTorch intent classifier, MLflow, and Docker Compose, plus an n8n workflow using GitHub API, Serper, and OpenAI GPT to catalog AI use cases in Google Sheets.
I built quantum rag evaluation as a Quantum Computing PDF question-answering project with MMR retrieval, FAISS, Groq Llama 3.1, and an automated evaluation framework. Its example evaluation reports semantic similarity 0.855, keyword coverage 0.154, BLEU-1 0.812, and PASS; these are example results, not broader performance measurements.
My computer-vision and language projects include vehicle detection with Python, OpenCV, and Roboflow API; receipt OCR and expense extraction with OpenCV and Tesseract OCR; and an IMDb sentiment-analysis project using Streamlit and machine learning. My multimodal safety assistant combines BLIP computer vision with Groq Llama 3 to analyze images, interpret ambiguous queries, and infer hazards such as electrocution, fire, and slip risks.
I document the implementation of my multimodal safety assistant in multi modal ai safety assistant: it uses BLIP image captioning and Groq Llama 3.3 to produce plain-language hazard reports.
My analytics repositories cover Excel sales and profitability analysis with KPI reporting, PivotTables, lookup formulas, and revenue and profit analysis; PostgreSQL analysis of Netflix and retail data using cleaning, EDA, business-intelligence queries, window functions, and trend analysis; and an Amazon sales notebook and Indian food analysis. The profile also lists repositories with limited descriptions, including Rag_pipline, SmartAssist-RAG, Sales-Agent, and Retail_sales-sql-rpoject. Sales-Agent has 1 fork. Listed repository update dates range from 2024-05 to 2026-10; the profile provides no model performance metrics or further detail about project users or deployment status.