Amit Kumar

Developed production-ready React-based landing page and dashboard components for data analytics SaaS platform, collaborating with design team to implement responsive UI/UX.
Key Achievements:

Built interactive analytics dashboard with real-time data visualization using React and Chart.js
Implemented responsive design LH Score 95
Developed reusable component library reducing development time by 30%

OmniDoc-AI

Built document intelligence system using hybrid retrieval (vector + keyword search) with multi-provider LLM integration that improved user comprehension by 45% and enabled 2x faster research document analysis.
Tech Stack: Python, LangChain, FAISS, React, UI/UX, OpenAI API, LLM's, LLamacpp, Ollama
GitHub: github.com/HmbleCreator/OmniDoc-AI

Modernized and optimized psychotherapy practice website, migrating from legacy Joomla CMS to contemporary web standards with integrated booking functionality.
Key Achievements:

Implemented integrated booking system reducing appointment scheduling time by 60%
Enhanced site interactivity and mobile responsiveness, decreasing bounce rate by 25%
Improved page load speed by 50% through optimization

AI-powered lead scoring & prioritization tool for SaaSquatch Leads. Analyzes leads across 5 dimensions, automatically tiers A/B/C/D, and provides actionable buying signals.
Tech Stack: Python, Gemini API, React, OpenAI API, PostgreSQL
GitHub: https://github.com/HmbleCreator/SaaSquatch-Intelligence-Hub

Led end-to-end architecture and product development for an AI-driven MVP, translating research concepts into production-ready prototypes [30% faster than projected / in an average of 2 weeks].

Architected a scalable AI/ML infrastructure that reduced data processing costs by 20% and enabled 3x faster / daily team iteration.

Solar Fault Detection

Engineered predictive maintenance system reducing false alarms by 40% using ensemble ML methods for proactive solar infrastructure monitoring.
Tech Stack: Python, Scikit-learn, Plotly
GitHub: github.com/HmbleCreator/solar-fault-detection-simulator

ScoreFI 2.0

Description: Developed end-to-end ML pipeline scoring 10K+ DeFi wallets with 35% reduction in false positives using behavioral feature engineering and graph-based Sybil detection.
Tech Stack: Python, Scikit-learn, Web3.py, PostgreSQL
GitHub: github.com/HmbleCreator/scoreFi-2.0

Rag- Chatbot

Created standalone RAG system reducing redundant LLM calls by 60% through semantic caching, running local GGUF models without external dependencies.
Tech Stack: Python, Flask, llama.cpp, GGUF Models
GitHub: github.com/HmbleCreator/rag-chatbot

Founded and scaled an open-source software community from 0 to 300+ active members in 3 semesters.

Organized 10+ technical workshops, hackathons, and mentorship programs focused on open-source contribution and best practices, resulting in 25+ new open-source Contributors.

Led a technical community of 300+ members, established standardized Git workflows for open-source contribution, and delivered workshops on Linux, Python scripting, and version control best practices.

Architected a multi-agent generative AI system using LangGraph and PyTorch to evaluate and optimize architectural floor plans, achieving 95% audit accuracy and supporting published IEEE research.

Built a production-ready RAG answering engine and data pipelines, handling data ingestion, embedding creation, and multi-provider LLM integration to optimize latency and cost for GenAI services.

Developed ML pipelines and graph-based Sybil detection for 10,000+ DeFi wallets, engineering 25+ features and reducing false positive risk assessments by 35%.

Grade: 9.15/10

Activities and societies: Research on Hybrid AI frameworks; coursework projects in deep learning and NLP.

Pursuing a B.Tech in Computer Science with a minor in AI/ML; coursework includes Deep Learning, Machine Learning, Algorithm Design, NLP, and Data Structures, with a research presentation accepted at IEEE IC3ECSBHI 2026.

Grade: 8.51

Activities and societies: FOSS CLUB(Technical And Mentoring Experience), Research Papers, Hackathons.

ML Engineer specializing in NLP and RAG architectures, with experience building 6+ AI applications including document intelligence systems and DeFi scoring models. Founded and scaled a 300-member open-source community, contributed to 7+ OSS projects, and recognized nationally as a top research candidate from 100K+ entries.