Ajinkya
Mane.
Final Year Computer Science student focused on AI, Machine Learning, LLMs, RAG Systems, and Full Stack Development. Building intelligent systems that bridge the gap between human needs and machine capability.
Who I Am
Hello there! I'm Ajinkya Mane, a dedicated Computer CS Engineer with a strong passion for developing robust and user-friendly applications.
My journey in tech began with a curiosity for how things work, quickly evolving into hands-on experience in both front-end development and intelligent systems (AI). I thrive in environments where I can leverage my skills to build innovative solutions.
I build AI systems that solve real business problems, not just proof-of-concepts. From architecting high-accuracy RAG (Retrieval-Augmented Generation) systems to implementing secure user flows, I aim to ship clean, optimizeable code.
Career History
Python Full Stack Developer Intern
ItVedant Pvt. Ltd.
- Built and containerized full-stack applications leveraging Django, Flask, React, and PostgreSQL.
- Optimized relational database queries, achieving a 30% speedup in average query response times.
- Engineered secure authentication systems (JWT, OAuth) and comprehensive admin command panels.
- Collaborated in Agile sprints, taking part in code reviews to ensure code quality standards.
Software Developer Intern
Infeanet Digital Marketing & Web Media
- Developed interactive web pages and UI components using modern frontend technologies.
- Integrated backend endpoints and managed client-side data state across multiple views.
- Assisted in debugging legacy code, refactoring utility classes to clean standard practices.
Projects Portfolios
LuminAI — AI Career Mentor
A unified portal providing ATS resume analysis, real-time AI mock interviews, custom roadmap generators, and chat mentoring.
ContextBot — RAG Document Q&A
A production-grade RAG pipeline using ChromaDB vector store, Hugging Face embeddings, and Groq Llama 3.3 for high-accuracy semantic search.
FitGenie — AI Nutrition & Fitness
An automated assistant that generates meal plans and 7-day workouts, analyzing plate photos using vision models.
XAI Model Monitoring Platform
An MLOps dashboard showing real-time data drift alerts, feature contribution scores, and local model explanations.
Alfred — Voice Assistant
A J.A.R.V.I.S.-inspired virtual assistant capable of executing system tasks and custom voice actions.
Multi-Device Home Management
Cloud-based home management featuring real-time database sync and multi-device command routing.
NGO Site Management
Dual-module NGO platform featuring a secure Admin Dashboard to edit and broadcast live organizational updates.
User Auth using Streamlit
A drop-in streamlit session validator featuring secure registration and login schemas.
PennyWise — Expense Tracker
Expense application featuring real-time balance sheets, categories, and automated budget threshold overrides.
Moodify — Mood Matcher
An API-integrated web application suggesting tracks matching real-time user sentiments and activities.
Technical Capabilities
Languages
AI & ML
LLMs & Retrieval
Frontend Stack
Backend Services
Databases
Tools & DevOps
Currently Exploring
Agentic AI
Constructing multi-agent networks that plan, evaluate, and self-correct tasks asynchronously.
LangGraph
Building stateful multi-actor systems with cyclic graph architectures for agentic workflows.
Explainable AI (XAI)
Applying SHAP, LIME, and integrated gradients to interpret neural net and ensemble decisions.
MLOps Pipelines
Automating model testing, data drift monitoring, and rolling deployments using Docker/CI.
Deep Learning
Experimenting with visual encoders, sequence models, and custom parameter-efficient fine-tuning.
Active Contributions
Profile Metrics
Top Languages
Featured Repositories
LuminAI-AI-Career-Mentor
Full-stack AI Career Mentor featuring ATS resume check, roadmap generator, and mock interviews using Gemini API.
ContextBot-RAG
Production-grade document Q&A system using ChromaDB vector store, Hugging Face, and Groq Llama 3.3.
FitGenie-AI-Fitness
AI-based Vision fitness assistant providing meal macronutrient estimation and customized workout generators.
Achievements & Certifications
ContextBot RAG Excellence
Architected a custom retrieval-augmented pipeline achieving 92% semantic accuracy while optimizing memory footprint by 40% using vector compression.
LLM Scale Integration
Integrated state-of-the-art APIs (Groq Llama 3.3, OpenAI) across 5 core applications, serving 500+ custom fitness schedules.
Cinema Booking System Publication
Co-authored and published comprehensive research covering system architecture and database load-balancing in IJRPR & IJIRCEE journals.
Let's Connect
Open to opportunities, collaboration proposals, and engineering dialogue. Let's build something serious.