I'm someone who is always curious about how things work - and even more curious about what I can create with them.
I'm an AI & Data Science graduate who enjoys turning curiosity into ideas, ideas into projects, and challenges into practical solutions. I explore data, experiment with AI, and learn by building - one idea at a time.
This portfolio is a glimpse into that journey - the things I've built, the ideas I've explored, and the lessons I've picked up along the way.
My journey has been shaped by curiosity, experimentation, and a constant desire to build.
From data analysis and interactive dashboards to AI applications, machine learning, RAG workflows, and digital experiences, I've explored different sides of technology while learning how ideas can become practical solutions.
Every project has challenged me to think differently, solve problems creatively, and keep learning along the way.
This is a glimpse into the work, ideas, and experiences that continue to shape how I think and build.
To take on meaningful challenges, keep learning, and contribute ideas that create real value.
To grow through new experiences, think beyond the obvious, and make a meaningful impact.
Most importantly, to keep building things worth building.
Hover the constellation to inspect a category — click a satellite to explore its stack.
LLM orchestration, prompt engineering, retrieval-augmented generation, and ML / vision / NLP libraries.
Core languages and web frameworks used across data and web applications.
Reporting, dashboarding, and visualization for data cleaning, analysis, and business insights.
Prototyping, wireframing, and layout planning with Figma and Canva.
Version control and the development environment used daily.
Developed a real-time AI food concierge with Streamlit, LangChain, and RAG for grounded menu retrieval, session summarization, and streaming analytics.
Real-time AI sentiment analyzer & document intelligence platform. Evaluates polarity, emotion drivers, and subjective tone across text, Excel, CSV, PDF, and Word documents with automated actionable recommendations.
Created three Power BI projects for Spotify, BigMart, and student analytics, covering data cleaning, analysis, visualization, KPI tracking, and actionable insights.
Built an AI-powered computer vision parking space detection platform using OpenCV, CNNs, and real-time edge density analysis to track occupancy, vacant spots, and live lot analytics.
Built an ML-based educational recommendation web app for personalized course recommendations using user profiles and interests.
A beautifully designed perfume e-commerce website showcasing products with an elegant UI/UX, smooth animations, and a seamless booking flow.
Premium portfolio and digital experience platform developed for a luxury branding studio, showcasing identity design, editorial brochures, packaging, and web development.
Published in the Institution of Engineers (India) Students' Chapter
Research paper on enhancing smart agriculture through digital twin technology.
Awarded by IEI-KKLC for outstanding research contributions in Artificial Intelligence & Data Science.
Avenora'24 National Level Technical Symposium.
Avenora'24 National Level Technical Symposium.
Biotronix'24, Rohini College of Engineering and Technology.
Axelyra v.25 National Level Technical Symposium.
Questions about my projects, internships or certifications? Send a message.
No forms, no delays — reach me instantly.
Drop me an Email
A real-time AI food concierge for grounded menu retrieval, session summarization, and streaming analytics.
Developed a real-time AI food concierge with Streamlit, LangChain, and RAG for grounded menu retrieval, session summarization, and streaming analytics.
Real-time sentiment analysis, multi-format document intelligence & emotion detection platform
Try text sentiment analysis, upload spreadsheets, or analyze PDFs in real-time.
Traditional sentiment analysis tools are limited to single short sentences. Organizations, researchers, and product teams struggle to process hundreds of customer reviews, bulk feedback spreadsheets (Excel, CSV), and multi-page reports (PDF, Word) in one place with actionable business guidance.
Engineered ZendiMind AI, a comprehensive document and text intelligence platform deployed on Vercel. It processes direct text input and batch files up to 32MB, extracting text columns from spreadsheets and documents, evaluating polarity, subjectivity, and emotion drivers, and generating instant recommendations and exportable reports.
Instant classification into Positive, Neutral, and Negative categories with fine-grained confidence scoring, polarity index, and subjectivity metrics.
Supports uploading Excel (.xlsx, .xls), CSV, PDF, Word (.docx), and Text files. Features automatic column detection and batch review parsing for thousands of records.
Detects emotional drivers including Joy, Sadness, Anger, Fear, and Neutral tone to understand customer intent and satisfaction nuance.
Delivers automated, context-aware improvement recommendations based on detected negative patterns to assist product teams in boosting customer retention.
Custom glassmorphic 3D robot preloader with audio soundscape toggle, particle celebration canvas, and instant light/dark theme switching.
Visual distribution charts, interactive review inspection tables, and 1-click export of analyzed datasets directly to CSV.
End-to-end music intelligence system analyzing 5,204 tracks, 2,482 artists, popularity distribution, explicit vs. clean ratios, and historical music trends spanning 1960 to 2020.
End-to-end retail sales intelligence system evaluating 1,449 items across multiple outlet establishment tiers, analyzing ₹4,33,972.15 in total item MRP revenue, outlet formats, and product visibility correlations.
Comprehensive Power BI analytics tracking student performance, enrollments, ₹6,877.50K revenue, 504 certificates issued, course trends, demographics, and actionable institution KPIs across 2 dedicated dashboards.
Interactive dashboards covering Spotify, BigMart, and student analytics data through cleaning, analysis, visualization, and business insights.
Built interactive Power BI dashboards for Spotify, BigMart, and student analytics, focusing on data cleaning, analysis, visualization, and business insights.
A real-time system to detect parking spaces and identify vacant spots.
Built a real-time parking space detection system using OpenCV to identify vacant spots.
An ML-based educational recommendation web app for personalized course recommendations.
Built an ML-based educational recommendation web app for personalized course recommendations using user profiles and interests.
Designing a visually appealing e-commerce platform that captures the essence of luxury fragrances while maintaining an intuitive booking process.
Built a responsive perfume booking shop from scratch using HTML, CSS, and JS, featuring hover effects, product showcases, and form validation.
Creating a digital presence that reflects the high-end, premium quality of a professional graphic design studio.
Developed a visually striking, animation-rich website with highly attractive UI elements and seamless navigation to serve as the ultimate studio portfolio.
Track performance, enrollments, revenue & certificates across course categories
Davinci Resolve · ₹1M Revenue
Leading all creative courses in student fee contribution followed by Adobe After Effects (₹1,188K) and Autodesk 3ds Max (₹1,175K).
November Peak · 116 Students
Enrollment surge observed in Q4 with steady intake peaks in April (91 students) and August (85 students).
20–29 Age Group · 52.07% (402)
Core learner base is young adults (52.07%), with 33.68% (260) in the 30–39 demographic seeking upskilling.
Average Rating · 4.3 / 5.0
High course satisfaction across 8 specialized courses with 65.3% certificate completion achievement.
Analyzing music trends, artist popularity & track distribution across 6 decades (1960–2020)
Leader · Taylor Swift (>4K Score)
Leading both total catalog tracks and overall popularity, followed by The Weeknd, Drake, Lil Peep, Hans Zimmer, Ariana Grande, and Lady Gaga.
Historical Peak · 1970–1980 Era
Longest average song durations peaked during classic album eras before steadily trending toward compact streaming-era durations (~3.2 mins).
Explicit Ratio · 23.31% (1.21K Tracks)
Tracked cross-sectional popularity across studio albums, compilations, and singles, highlighting evolving listener consumption habits.
Retail performance intelligence, outlet tier breakdown & item visibility analysis across grocery channels
Dominance · Supermarket Type 1 (80.07%)
Supermarket Type 1 generates ₹347.46K of sales while Supermarket Type 2 contributes ₹86.51K (19.93%), highlighting the immense value of standard supermarket footprints.
Top Categories · Snack Foods (63K) & Fruits (61K)
High-frequency consumables lead grocery sales volume, followed by Household items (51K), Frozen Foods (39K), Dairy (34K), and Canned goods (32K).
Revenue Flow · Tier 1 (₹1,74,486.80) Lead
Root sum of ₹4,33,972.15 decomposes directly into store format and geographic location tiers, proving Tier 1 urban locations yield double the volume of Tier 3 (₹87,363.43).
