About
I am Alby Ponnachan, a Computer Science undergraduate passionate about AI/ML and Full-Stack Development. I enjoy building intelligent systems and modern web applications that solve real-world problems and create meaningful user experiences. I have hands-on experience with Python, TensorFlow, PyTorch, React, Flutter, and Flask, along with tools like Hugging Face, OpenAI APIs, n8n, and Git. My work focuses on combining artificial intelligence with scalable web and mobile technologies. I have developed a range of projects including a Smart Refrigerator App, an AI-based phishing detection system, a malware detection model, an LLM-powered disease query system, and a personalized finance tracker with AI insights. I have also built full-stack applications like a Flask blog platform and the YASYA’26 fest website, along with automation workflows using APIs. I am always curious to learn new technologies, improve my skills, and build impactful solutions at the intersection of AI and software development.
Experience
-
Cybersecurity InternTechbyheart2025 – 2025
-
Web Development InternKeltron2023 – 2023
Education
-
B.TechSt. Thomas Institute for Science and Technology · APJ Abdul Kalam Technological UniversityComputer Science and Engineering · 2022 – 2026
-
QualificationDiploma in Computerised Financial Accounting · KELTRONAccounting · 2020 – 2021
-
High SchoolSt. Johns Higher Secondary School2018 – 2020
Skills
Projects
-
React, Groq API, PubMed API, PDF.js
Full-stack React SPA utilizing the Groq API for streaming clinical Q&A, NCBI PubMed API for peer-reviewed literature search, and client-side PDF.js for secure, backend-free medical report analysis.
-
Next.js 14, LangGraph, RAG, Firebase, Gemini, Groq
Agentic finance manager featuring an autonomous budget-monitoring agent, a local RAG pipeline for receipt analysis via Google Gemini, and real-time financial consulting powered by Llama 3.3 via Groq.
-
Python, TensorFlow, Keras, Gradio, Hugging Face
Medical imaging system using ResNet50 transfer learning to detect pneumonia from chest X-rays. Implemented Grad-CAM explainability to visualize decision regions, achieving 97.37% accuracy and 99% recall, deployed as an interactive Gradio app on Hugging Face Spaces.
-
LangChain, Groq (LLaMA 3), RAG, Python
Intelligent document processing system using RAG for automated PDF metadata extraction and natural language question-answering over large document sets.
-
Python, TensorFlow, MediaPipe, OpenCV, Gradio, gTTS
Built a real-time sign language recognition system using MediaPipe hand landmark extraction and an LSTM neural network for temporal gesture classification. Integrated confidence filtering, majority voting, and anti-hallucination logic for stable predictions, with Text-to-Speech output and live deployment on Hugging Face Spaces.
-
LangGraph, Supabase, Groq Llama 3.1, Next.js 15, React 19, TypeScript, FastAPI
Architected and deployed a production-ready multi-agent AI SaaS platform where six specialized AI agents collaborate using LangGraph workflows with RAG, parallel execution, and automated security review. Built a secure FastAPI backend with JWT authentication, multi-tenant architecture, real-time monitoring, file analysis, and automated Markdown/PDF report generation, deploying the application using Vercel, Render, and Supabase.
Courses & certifications
- Artificial Intelligence A–Z 2025 · Udemy · 2025
- Google Project Management · Coursera
- The Complete Full-Stack Web Development Bootcamp · Udemy
- AI Automation: Build LLM Apps and AI Agents with n8n · Udemy
- Machine Learning with Python · Coursera
- Deep Learning with Keras and TensorFlow · Coursera
- Introduction to Neural Networks and PyTorch · Coursera
- Fundamentals of AI Agents Using RAG and LangChain · Coursera
- Gen AI Foundational Models for NLP & Language Understanding · Coursera
- Generative AI Engineering and Fine-Tuning Transformers · Coursera
⚽ Extracurricular activities
-
Tech Lead
Led the technical team for the college cultural fest and developed the official fest website.
-
Basketball Team Leader
Led the squad to victory in the annual inter-college sports event.
-
Community Engagement
Participated in multiple national hackathons and advanced technical workshops.