RESUME
Software Developer
Link to the Resume: Here
As a software developer, I thrive on solving complex problems through automation, AI, and full-stack development. My adaptability has allowed me to work across enterprise systems, research, and startups, enabling me to deliver impactful solutions in diverse environments.
AI / ML Researcher
Link to the Resume: Here
As an AI/ML researcher, I am driven by curiosity and a passion for building real-world solutions. I adapt quickly to new models and platforms, from optimizing ADAS detection with YOLOv8 to streamlining Generative AI pipelines, ensuring impactful contributions to applied AI and machine learning.
PROFILE
Co-Founder
Co-founder of Trinite, driving innovative tech solutions to help small businesses thrive. Link to Trinite's website: Here
Developer
Developer with 3+ years of experience creating AI-driven, automated, and full-stack solutions across industries.
Researcher
AI/ML researcher with experience advancing ADAS detection, Generative AI, and large-scale deep learning models.
Student
MSc Computer Science graduate from Western University, specialized in Artificial Intelligence.
About Me
I am a MSc Computer Science graduate (Artificial Intelligence) from Western University with over 3 years of experience spanning AI/ML research, enterprise systems, and full-stack development. My journey has taken me from co-founding startups to contributing at large organizations. At London Hydro, I optimized 35+ SAP CDS Views, cutting runtimes from 30 minutes to 300 ms, and automated XML/PDF pipelines that streamlined billing validation. During my Generative AI Externship at Cognizant, I delivered 30+ GenAI projects and built Azure ML pipelines that reduced deployment time by 40%. Alongside, my research on Advanced Driver Assistance Systems (ADAS) explored YOLOv8 and RT-DETR models, training 70+ custom models and publishing a 100-page thesis. I thrive at the intersection of AI innovation, enterprise engineering, and scalable software development.
My Coding Journey
My passion for programming began in high school with Java, which sparked an early curiosity for problem-solving. In my undergrad, I expanded into C and Python through IBM’s AI/ML program, laying the foundation for my future in artificial intelligence.
As I grew my skills, I ventured into web development, mastering.........
Read More
SKILLS
Core Stack
Languages & Tools
Frameworks & Libraries
Specialized Expertise
EDUCATION
Hiranandani Foundation School, Thane
ICSE 10th - 2017
93.4% (Best of five: 94%)
Shubhamraje Junior College, Thane
HSC 12th - 2019
81.54%
Joint Entrance Examination (Main)
2019
National Level engineering entrance assessment test.
Score: 96.93 percentile
Maharashtra Common Entrance Test (MHT-CET)
2019
State Level engineering entrance assessment test.
Score: 98.78 percentile
Dwarkadas J. Sanghvi College of Engineering (Affiliated to Mumbai University)
2019 - 2023
B.Tech - Computer Engineering
Pointer: 9.58/10 (or 89.92%)
IBM & Dwarkadas J. Sanghvi College of Engineering
2019 - 2023
Minor Specialization - Artificial Intelligence and Machine Learning
Pointer: 9.32/10 (or 87.01%)
University of Western Ontario
2023 - 2025
M.Sc - Computer Science (Artificial Intelligence)
GPA: 4.00/4.00 (or 94.5%)
PROFICIENCY AND APTITUDE TESTS
Graduate Record Examinations (GRE) - General
11th November, 2021
Score: 319/340
---- Quantitative: 168/170
---- Verbal: 151/170
---- AWA: 4/6
Canadian English Language Proficiency Index Program (CELPIP) - General
13th February, 2025
Average: 11
---- Listening: 11
---- Reading: 11
---- Writing: 11
---- Speaking: 9
EXPERIENCE
London Hydro
Utility Information Technology Intern01st May 2025 - 31st Aug 2025
At London Hydro, I worked across SAP S/4HANA and internal automation. I built 35+ SAP CDS Views (vendor scope 14) and optimized critical reports from ~30 minutes to ~300 ms by converting legacy SQL and aligning mappings with business/technical teams. On the QA side, I reduced OCM defect tickets from 256 to 0 by standardizing 50+ guides and improving UAT tracking (100+ JIRA tickets). I also automated XML/PDF pipelines (Python, Flask, Spring Boot): bulk validation of 24+ Green Button files, SOAP-based bill fetching, and 4-way invoice consistency checks integrated with WatchDog—accelerating billing validation and enterprise reporting.
Cognizant
Externship Generative AI Extern01st Jan 2025 - 31st Mar 2025
Certificate
During a Generative AI externship, I delivered 33+ real-world GenAI projects spanning prompt engineering, retrieval, and LLM fine-tuning. I streamlined deployment on Azure ML with MLOps pipelines that cut time-to-production by ~40%, and standardized evaluation/monitoring for reliability. Work included building production-ready LLM workflows (chat, summarization, classification), integrating vector search, and optimizing inference cost/latency. I earned the Cognizant GenAI Digital Badge for demonstrating proficiency in enterprise-scale GenAI solutions.
University of Western Ontario
University Graduate Research Assistant and Graduate Teaching Assistant01st Sep 2023 - Present
As a Graduate Teaching Assistant for CS1026 and CS1027 at Western University, I taught core computing in Python and Java, led weekly labs and office hours, and pre-tested assignments to ensure clarity and correctness. In parallel, as a Graduate Research Assistant in applied computer vision, I investigated real-time object detection for ADAS, comparing YOLOv8 and RT-DETR. I trained and evaluated 35 custom models with PyTorch on Compute Canada GPUs/Google Colab, analyzing inference speed and accuracy; this work showed that YOLOv8 outperforms RT-DETR in ADAS settings and culminated in a 100+ page thesis and a comprehensive paper. This dual role strengthened my teaching, research, and communication skills—bridging rigorous ML experimentation with clear, student-focused explanations.
Kwoda
Start-up Software Engineer Intern01st Sep 2023 - 01st Jul 2024
Website Link
Certificate
As a Software Engineer Intern at Kwoda, I built a multi-company HR platform with Django/DRF and React on MySQL, streamlining onboarding and performance workflows for 100+ employees across 3 companies. I engineered a verified digital-resume system that generates tamper-proof records from manager-approved data and added QR-based instant verification, driving 100% adoption and strengthening employer trust. I designed REST APIs, implemented approval pipelines, and optimized database queries to ensure reliable, low-latency operations. The result was faster HR cycles, greater transparency, and a scalable foundation for future modules.
Trinite
Start-up Co-founder and Full-Stack Developer15th Jun 2021 - 01st Sep 2023
Website Link
Certificate
As Co-Founder & Full-Stack Developer at Trinite, I delivered custom web solutions for 5+ small businesses, focusing on scalable builds and measurable growth. A highlight was Knickgasm—an e-commerce platform built end-to-end with Django/DRF, React, and MySQL, featuring secure checkout via Razorpay, user auth, cart, and wishlist. Through rigorous unit/functional testing, the site achieved 100% reliability and processed 600+ transactions post-launch, contributing to 15+ celebrity collaborations and a Shark Tank India pitch. I owned backend APIs, database design, and CI with Git, ensuring performant, maintainable releases that helped clients strengthen their online presence and operations.
Threco (The Recycling Company)
Internship Software Developer Intern01st Jun 2021 - 17th Aug 2021
Website Link
Certificate
As a Software Developer Intern at THRECO, I engineered dual portals (B2B & B2C) for waste management that enabled real-time transaction tracking, location visibility, and automated certificate generation for 60+ businesses and 100+ residential societies. I designed Django/DRF APIs with React front-ends and SQLite, implemented workflow automations that eliminated manual steps and improved data accuracy, and delivered features that increased transparency and reduced turnaround time. The result was a scalable platform that streamlined operations and boosted client satisfaction while supporting THRECO’s sustainability mission.
DJ ACM
Student Committee Full-Stack Developer Intern01st Sep 2020 - 01st Jun 2021
Website Link
Certificate
As a Full-Stack Developer Intern at DJ ACM (ACM student chapter), I built a mobile-friendly site to boost engagement and automate event workflows. I delivered dynamic pages for event showcases, chapter history, and member highlights that increased user engagement by ~35%. I integrated Razorpay for seamless online registrations and payments, facilitating collections of INR 15,000+ for flagship events (e.g., “Lines of Code,” “Internship Fair”) and significantly reducing manual effort. The stack included Python, Django, HTML/CSS/JavaScript with Bootstrap for responsiveness and Git for version control—delivering a robust, user-friendly platform that streamlined operations for organizers and participants.
DJ UNICODE
Student Community Full-Stack Developer Mentee01st Sep 2020 - 15th Aug 2021
Backend Django Mentor
16th Aug 2021 - 18th Jul 2022
BE Mentor - Treasurer
19th Jul 2022 - 01st Jul 2023
Website Link
Certificate
As a Full-Stack Developer Mentee (Aug 2020 – Jul 2021), I built 6 client web apps, contributing across Django/DRF backends and React, HTML, CSS, JavaScript frontends. I shipped features end-to-end (APIs, DB models, UI) and learned to deliver scalable, responsive solutions through close collaboration and code reviews.
Promoted to Backend Developer Mentor (Jul 2021 – Jul 2022), I led teams on 3 projects—including a JAAV charity site that increased donations by ~30% and a Student Portal reaching ~70% active engagement. I designed and deployed REST APIs, guided debugging/testing, and mentored juniors in Django/DRF and API design to production standards.
As Treasurer & Mentor (Jul 2022 – Jul 2023), I managed finances and ops, implementing cost controls with Excel/Google Sheets that reduced organizational spend by ~15%. I coached members on project execution and budgeting, and streamlined cross-team collaboration using Slack, strengthening the chapter’s long-term sustainability.
Phemesoft
Internship Backend Developer18th Jul 2022 - 12th Sep 2022
Website Link
Certificate
As a Backend Python Developer at Phemesoft, I enhanced the Virtual Workspace collaboration platform by designing and shipping Django REST Framework APIs, optimizing PostgreSQL queries/indexes, and refactoring backend services for reliability and speed. These improvements increased platform efficiency by ~20%, reduced response times, and stabilized data flows. I collaborated via Git-based workflows, implemented clean API contracts for seamless front-end integration, and strengthened observability to ensure a responsive, high-performing experience for users.
JPMorgan Chase & Co
Technology Analyst Intern - Industry Skills Development Program Student11th Oct 2020 - 25th Sep 2021
Certificate
As a participant in JPMorgan Chase & Co.’s Industry Skills Development Program (Technology Analyst track), I built automation and analytics prototypes using Automation Anywhere (RPA) and Alteryx. I designed bots to replace multi-step manual tasks (data entry, reconciliations) and created ETL pipelines that cleaned, joined, and validated datasets for reporting, strengthening decisions with reliable data. I also practiced bank-grade SDLC—Git-based workflows, documentation, and reviews—and tuned SQL queries for faster reporting. The program sharpened my foundations in process automation and data engineering, preparing me for enterprise-scale, accuracy-critical environments.
PROJECTS
List of 23 important projects undertaken. Click on them to learn more
- All
- Company
- E-commerce
- Descriptive
- Community
- Machine-Learning
- Java
- Deep-Learning
- Cybersecurity
Trinite Website
Company, Descriptive
Knickgasm E-commerce
Company, E-commerce
Technofocus
Descriptive
JAAV Website
Company, Descriptive, Community
DJ ACM Website
Company, Descriptive, Community
Attendance using ML
Machine-Learning, Deep-Learning
DJ Events
Descriptive, Community
Student Portal
Community
Threco Dashboard
Company
NconNect
Descriptive, Community
Traffic Sign Classifier
Machine-Learning
Stress Detector Application
Machine-Learning
Hughes Holidays Website
Company, Descriptive
SND Logistics Website
Company, Descriptive
World22 Logistics Website
Company, Descriptive
Tritech Solutions Website
Company, Descriptive
DJ ICACTA Website
Descriptive, Community
Selective Encryption
Machine-Learning, Deep-Learning, Cybersecurity
Multilayer AV Cryptography
Cybersecurity
Kwoda HR Portal
Company
VisionForge (Stable Diffusion 1.5)
Machine-Learning, Deep-Learning
Vehicle Specific Detection
Machine-Learning, Deep-Learning
RESEARCH
Real-Time Performance Dynamics in ADAS: YOLOv8 vs RT-DETR
Co-Authors: Dr. Michael Bauer
Conducted a comparative study of YOLOv8 and RT-DETR for real-time ADAS detection, training and evaluating 35+ models on BDD100K and custom datasets. Benchmarked mAP, F1, and inference time; results showed YOLOv8 achieved faster latency and higher accuracy, handling class imbalance more robustly. These findings guide model selection for production-grade ADAS where milliseconds and precision matter.
Comparison of Tabular Synthetic Data Generation Techniques using Propensity and Cluster Log Metric
Co-Authors: Dr. Ramchandra Mangrulkar, Dr. Aruna Gawde, Mr. Aryan Pathare, Mr. Kartik Suvarna
Compared tabular synthetic data generators across multiple datasets, evaluating utility with Propensity Score and Cluster-Log metrics. Benchmarks quantify realism and structure preservation, outlining method-specific strengths. Manuscript submitted to the International Journal of Information Management Data Insights and currently under review.
Multilayer combined symmetric and asymmetric approach for audio and video cryptography
Co-Authors: Dr. Ramchandra Mangrulkar, Mr. Mayav Antani, Mr. Kartik Suvarna
Developed a multi-layer cryptography scheme combining symmetric and asymmetric techniques, deriving multiple session keys from a single private key. The method encrypts both audio and video streams in audiovisual files while remaining user-friendly and secure. The paper is in final edits and will be submitted for publication soon.
Real Time Multimedia Encryption
Co-Authors: Dr. Narendra Shekhokar, Mr. Mayav Antani
Designed a low-latency real-time multimedia encryption method for concurrent audio and video streams, targeting higher throughput and lower latency than baseline schemes. Proposes a lightweight cipher pipeline with efficient key management to maintain strong security under streaming constraints; currently in planning/prototype stage as a capstone project.
Stress Detection using Twitter Analysis and PSS
Co-Authors: Prof. Khushali Deulkar, Mr. Kartik Suvarna, Mr. Mayav Antani, Miss Karishni Mehta
Built an NLP pipeline to score user stress (1–5) from Twitter posts using text preprocessing, feature extraction (TF-IDF/embeddings), and supervised ML classifiers. Benchmarked models (e.g., Logistic Regression, SVM, Random Forest) and selected the best performer; validated predictions against the Perceived Stress Scale (PSS) to ensure alignment and robustness.
Comparison study and complexity analysis of different methods for Maximizing flow in a network using Ford Fulkerson
Co-Authors: Prof. Sudhir Bagul, Mr. Kartik Suvarna, Mr. Mayav Antani
Comparative analysis of augmenting-path strategies in Ford–Fulkerson: Edmonds–Karp (BFS), DFS-based, and a basic/greedy variant. We derive and validate complexity bounds (e.g., O(VE²) for Edmonds–Karp) and benchmark on synthetic graphs; results show Edmonds–Karp offers more predictable convergence/scalability, while DFS has higher variance and worst-case stalls—informing algorithm choice for maximum flow.