Errunds
On-Demand Delivery & Services App for Canada
Stack
A Swiggy-style on-demand delivery and services platform built for a Canadian client over a year-long engagement — featuring customer apps for Android and iOS alongside a full admin panel for managing vendors, orders, and operations.
Challenge
The client needed a complete multi-sided delivery ecosystem from scratch — customer ordering, vendor management, and live order operations — built to handle real-world delivery logistics across a Canadian market.
Outcome
Delivered native Android and iOS customer apps plus a comprehensive admin panel, enabling the client to onboard vendors, manage live orders, and run delivery operations end-to-end.
Key Features
ERP Suite
Inventory, Sales & HR Management Platform
Stack
A business ERP web application that unifies inventory management, sales order processing, and HR operations into a single dashboard — replacing disconnected spreadsheets and manual processes with one streamlined system.
Challenge
The client ran inventory, sales, and staff management across disconnected spreadsheets and tools, making reporting slow and error-prone with no single source of truth.
Outcome
Delivered a unified ERP dashboard covering inventory, sales, and HR with role-based access and real-time reporting, cutting manual reconciliation work and giving leadership a live view of operations.
Key Features
Clinic OS
Patient Appointments & Records Management
Stack
A clinic management system that handles patient appointment booking, doctor scheduling, and digital patient records — helping clinics move away from paper registers and phone-based bookings.
Challenge
The clinic managed appointments and patient records manually, leading to double-bookings, lost records, and long front-desk wait times.
Outcome
Delivered an appointment booking and patient records system with doctor scheduling, cutting front-desk workload and giving staff instant access to patient history.
Key Features
ZestPath
Indoor Navigation for Malls & Hospitals
Stack
An indoor wayfinding app that guides visitors through large indoor spaces — malls, hospitals, and campuses — with turn-by-turn directions where GPS alone doesn't work.
Challenge
Visitors inside large indoor facilities had no reliable way to find specific stores, departments, or rooms, leading to confusion and wasted time since GPS doesn't work indoors.
Outcome
Built an indoor navigation app with interactive floor maps and turn-by-turn directions, helping visitors find destinations quickly inside large buildings.
Key Features
Ledger
Smart Cash Book for Android & iOS
Stack
A cross-platform mobile application built for individuals and small businesses to track cash flow — log income, expenses, and maintain clean financial records on the go with offline support and real-time sync.
Challenge
Small businesses struggled with error-prone paper-based bookkeeping. The client needed a zero-learning-curve app that worked offline and synced in real time across devices.
Outcome
Delivered a clean, intuitive app with real-time Firebase sync, offline support, and PDF export — helping hundreds of users maintain accurate financial records with no accounting knowledge required.
Key Features
CMMS
Computerized Maintenance Management System
Stack
Enterprise-grade web software for large manufacturing companies to manage machine breakdowns, schedule preventive maintenance, assign work orders to technicians, and track full repair histories — replacing manual spreadsheet processes.
Challenge
The client's team tracked machine breakdowns in spreadsheets and paper logs, causing delays, missed maintenance windows, and difficulty auditing past repairs.
Outcome
Delivered a robust enterprise application with role-based dashboards, automated scheduling, real-time alerts, and a full audit trail. Maintenance response times dropped significantly.
Key Features
LaTeX AI Builder
Live Equation Builder Powered by AI
Stack
A web application that lets researchers and academics write, render, and publish complex mathematical equations in real time using LaTeX syntax — with an AI layer for natural language to LaTeX conversion and smart auto-completion.
Challenge
Writing LaTeX is tedious even for experts. The client needed a tool that let users describe an equation in plain English and get correct LaTeX output instantly with live preview.
Outcome
Built a real-time editor with live rendering, AI-powered natural language input, equation templates, and export to LaTeX, PNG, and PDF. Adopted by university departments for academic workflows.
Key Features
EduPath
Smart Learning Platform for Students
Stack
A comprehensive mobile learning platform that delivers personalised course content, quizzes, and progress tracking to students of all ages. Teachers can create and publish lessons, assign homework, and monitor class performance in real time.
Challenge
Educational institutions needed a scalable, engaging digital platform to supplement classroom teaching — one that worked on low-end Android devices and in areas with limited internet connectivity.
Outcome
Launched a full-featured app with video lessons, interactive quizzes, live class sessions, and offline content download. Student engagement and test scores improved measurably across pilot schools.
Key Features
ParkSense
Parkinson's Disease Detection via Deep Learning
Stack
A deep learning system that analyses hand tremor patterns, gait data, and voice recordings to detect early-stage Parkinson's disease — enabling clinicians to make faster, data-driven diagnoses before symptoms become severe.
Challenge
Parkinson's is often diagnosed late due to its subtle early symptoms. Neurologists needed an AI-assisted tool to analyse multimodal patient data and flag high-risk individuals earlier in the disease progression.
Outcome
Achieved 94% diagnostic accuracy on validation datasets using an ensemble of CNN and LSTM models. The system integrates directly into clinical workflows via a simple web interface.
Key Features
AutiScan
Autism Spectrum Detection using DL & AI
Stack
An AI-powered diagnostic aid that analyses behavioural video observations, questionnaire data, and eye-tracking patterns to assist clinicians in early autism spectrum disorder (ASD) detection in children — significantly reducing assessment time.
Challenge
ASD diagnosis currently relies on lengthy clinical observation and subjective expert assessment. Families often wait months for a diagnosis. The goal was to build an objective, data-driven screening tool accessible to general practitioners.
Outcome
Developed a multi-modal AI system combining video analysis, NLP-based questionnaire processing, and eye-gaze tracking to produce a risk-stratified ASD screening report with 91% sensitivity.
Key Features
RouteRx
Hospital Routing & Patient Flow Web Platform
Stack
A real-time hospital routing and patient flow management system built for Canadian healthcare networks. The platform helps patients find the nearest available emergency room, tracks ER wait times, and intelligently routes ambulances to reduce congestion across hospital networks.
Challenge
Canadian emergency departments frequently experience overcrowding. Patients and EMS teams lacked a unified real-time system to know which hospitals had capacity — leading to dangerous delays and misdirected ambulances.
Outcome
Deployed across a regional hospital network with live ER capacity feeds, ambulance routing, and patient self-triage. Average ER wait times in pilot hospitals reduced by an estimated 22%.
Key Features
NeuroSight
Brain Tumor Detection via Deep Learning
Stack
A deep learning–powered MRI analysis tool that automatically detects, segments, and classifies brain tumors from MRI scans. Designed to assist radiologists by flagging high-priority scans and reducing reporting turnaround time in oncology departments.
Challenge
Radiologists face an overwhelming volume of MRI scans daily. Manual review is slow and subject to fatigue-related errors. The client needed an AI tool that could pre-screen scans and highlight anomalies before the radiologist reviews them.
Outcome
Trained a U-Net segmentation model and ResNet classifier on 10,000+ annotated MRI scans. Achieved 96.3% tumor detection accuracy. Integrated into PACS radiology workflow as a secondary reader.
Key Features
RetinaAI
Diabetic Retinopathy Screening with AI
Stack
An AI-based fundus image analysis platform that screens patients for diabetic retinopathy — the leading cause of preventable blindness. The system grades retinal images automatically and flags urgent cases for ophthalmologist review.
Challenge
Millions of diabetic patients require annual eye screenings, but ophthalmologist capacity is limited. Many patients in rural and underserved areas never receive a timely screening, leading to avoidable vision loss.
Outcome
Deployed in primary care clinics and diabetic centres, enabling non-specialist staff to capture retinal images and receive an AI grading report within seconds. Sensitivity of 97% for referable retinopathy.
Key Features
Got a Project in Mind?
Let's talk about what you want to build. We'll take it from idea to launch.
Start a Conversation