Our Work
Every project we've shipped — websites, AI agents and automations, products, and custom software — with the real metrics behind each one.
Attendify
Attendify is a privacy first membership management app for gyms, sports academies, and tuition centers, handling QR attendance, fee reminders, and income tracking in one dashboard. Built for speed, it loads in under a second and stays reliable at scale.
Author Portfolio
A portfolio website built with Next.js and Tailwind to showcase a published author’s books and route readers straight to Amazon and Flipkart. It hit a perfect Lighthouse SEO score and lifted book sales through strong, trust building testimonials.
Avittam Homes Website
A full redesign of a real estate developer’s website, modernizing the interface and tuning performance and SEO without touching the existing stack. The result is a faster, more premium browsing experience that turns more visitors into leads.
KMJ & Co.
A professional online presence for a chartered accountancy firm, built to rank locally and capture enquiries automatically through WhatsApp and Google Sheets. The site hit perfect Lighthouse scores for SEO and best practices, with zero ongoing software cost.
MedicalData Project
A confidential healthcare AI initiative involving the development of machine learning models and LLM powered insights for predictive analytics on sensitive medical datasets. Extensive data cleaning and preprocessing ensured model reliability, with a strong focus on minimizing false positives and false negatives. Hosted securely on a local server to comply with strict data privacy standards.
Hazlo Project
A confidential cross platform mobile application developed using React Native and Firebase, built as an MVP to validate market fit for a sports industry solution. Optimized for production deployment on both Android and iOS, the app delivers a smooth user experience, intuitive navigation, and reliable backend performance to support initial user testing and feedback cycles.
FlintM
Developed the official website for FlintM, a marketing agency, with a strong emphasis on brand storytelling and visual impact. Delivered a custom video led hero creative designed to instantly engage visitors, alongside a sleek, high performance interface. With Lighthouse scores above 90% in SEO and performance and 100% in best practices, the site effectively showcases client results, ROAS, and case studies, building trust and credibility.
The MUN Company
Revamped the official website for a Model United Nations (MUN) company, enhancing its professionalism and appeal to schools and MUN enthusiasts. Built on WordPress, the redesign included high quality Canva designed visuals to strengthen brand identity and improve event promotion. Achieved 100% performance, 90%+ SEO, and top scores in best practices and accessibility.
Myndz Badminton Academy
Developed a dynamic portfolio website for Myndz Badminton Academy, showcasing star players at national and international levels, coach profiles, parent testimonials, and the academy’s values. Integrated Instagram feed and media gallery to highlight activities and achievements, building trust and brand credibility.
Plant Species Classification
Developed an AI driven platform to detect and recognize indigenous medicinal plants of India, providing details on their medicinal value and culinary usage. Trained on a dataset of 2,626 plant species (including leaves, flowers, roots, stems, bark, and fruits), the system utilized ResNet50, MobileNetV2, and VGG16 CNN models, achieving 95% accuracy through stochastic weight averaging. Integrated GPT-3 with enriched data from web scraping to enable instant Q&A about plants.
Loan Approval System
A confidential fintech AI solution built to predict loan approval likelihood using historical datasets of approved and rejected applications. The system employed a supervised learning model to deliver 85% accuracy, with low false positives and false negatives. Integrated directly into the loan processing workflow, it provides a probability score, the model’s decision, and then routes the application for final human verification, streamlining decisions and reducing processing time.
StockMarket Automation
A confidential trading system that pairs rule based logic with machine learning to automate buy and sell decisions across market indicators. It cut manual trading effort dramatically, sped up decisions, and grew revenue for the client.
TT Sports Academy
Revamped the website for TT Sports Academy, a multi sport facility offering badminton, yoga, boxing, swimming, and more. Built with HTML, CSS, and JavaScript, the redesign focused on boosting SEO from ~55% to 100% and improving best practices to 100%, significantly enhancing brand visibility and online credibility. Performance scores improved from 40% to 73%, making the site faster, more discoverable, and more engaging for potential members.
Techiegen
Built a meme driven learning platform as a Progressive Web App (PWA) for Techiegen, enabling students to learn academic concepts through humor and community interaction. Powered by Firebase on the backend, the platform was launched both as a PWA and on the Play Store, allowing rapid go to market testing for product validation and monetization potential. Optimized for high performance, accessibility, and SEO to maximize reach and engagement.
Temple Billing System
A desktop billing and ticket management system for a temple, replacing paper records with stock tracking, revenue dashboards, and digital receipts. Built for speed, it loads almost instantly and keeps every transaction accounted for.
Data Analysis Tool
Built a no code data analysis and predictive modeling platform for users with strong statistical knowledge but no programming experience. The tool allows users to upload CSV datasets, generate interactive charts and graphs, and build basic machine learning models such as linear regression and logistic regression directly through an easy to use interface. Developed with React and Flask, the platform streamlines data exploration and modeling for rapid insights.
ColorCoding Tool
Developed a computer vision chemical analysis tool to predict methanol and ethanol percentages in solutions using image uploads. Designed for professionals in chemistry, food safety, and quality control, the tool replaced expensive lab equipment with an affordable, fast, and highly accurate alternative. Achieved 93%+ prediction accuracy with results generated in under 3 seconds, making it a practical solution for real world industrial applications.