Jaineel ModiToronto, OntarioGraduating May 2028

Software
developer who
ships things.

I'm on RBC's mortgage approval platform team writing Java and Spring Boot APIs. On my own time I maintain AudioMark, a practice player for dancers, and publish the pieces I pull out of it.

Jaineel Modi
Queen's University, Bachelor of Computer Science (AI)

About 5,000 people practise to something I built.

AudioMark on Google Play

Selected work

06 projects
  1. 01

    Co-built a practice player for dancers. Loop a section, slow it down, and both settings save per track. 5,000+ active users on Google Play.

    Flutter · Dart · just_audio · iOS / Android

  2. 02

    section_loop

    A-B section looping for any Dart audio player, published on pub.dev and extracted from AudioMark. Handles the seek storms that make looping hard to get right.

    Dart · Flutter · pub.dev

  3. 03

    Tossit 2.0

    First place at SatecHack. Photograph an item and a VGG16 classifier sorts it into the right recycling bin, served through ONNX and FastAPI.

    Next.js · TypeScript · FastAPI · ONNX · VGG16

  4. 04

    MovieMatchApp

    Swipe-based recommendation engine that learns taste from like and dislike signals. Watchlists sync in real time.

    SwiftUI · Firebase · Flask · PostgreSQL

  5. 05

    RecSend CLI

    API testing from one YAML file. Declare a request and the response you expect, get a pass/fail that exits non-zero for CI.

    Python · Click · YAML · CI/CD

  6. 06

    MiniLM Fine-Tuned

    Fine-tuned all-MiniLM-L6-v2 on SNLI and MNLI with an NLI softmax loss, reaching 91.2% validation accuracy.

    Python · PyTorch · Transformers

About

A volleyball player, a ramen connoisseur, and a programmer.

I'm studying Computer Science with an AI specialization at Queen's University, graduating May 2028. Before RBC I traced failures in multi-agent LLM coding systems at the RISE Lab, and built a GPT-4 Revit copilot at Mulvey & Banani that saved 26 hours per project. The work I'm proudest of is AudioMark, which I co-built and still maintain.

Lately I've been contributing upstream: two documentation fixes to Next.js, found by auditing every error link the framework hands its users against the pages that actually exist.

Experience6 roles

Shopify

Incoming Winter 2027

Software Developer Intern

Royal Bank of Canada (RBC)

May 2026 – Dec 2026

Software Developer Intern

  • Client-facing REST APIs for the mortgage approval platform, spanning 10,000 applications.
  • A Spring Boot service that validates mortgage documents against business rules, cutting manual effort 95%.

Java · Spring Boot · REST APIs · MongoDB

Queen's University (RISE Lab)

Sept 2025 – Jan 2026

Student Researcher

  • Traced failures in multi-agent LLM coding systems to the agent and step that break a repair.
  • Built the repair pipeline in PyTorch, lifting Qwen2.5-Coder pass@1 on HumanEval by 12.8 points.

Python · PyTorch · LLMs · Research

Mulvey & Banani International

Summers 2024 & 2025

Software Developer Intern

  • Shipped a GPT-4 Revit copilot automating 20+ BIM tasks for 80+ users.
  • Automated blueprint generation with the Revit API and C#, saving 26 hours per project.

GPT-4 · Python · C# / .NET · Revit API

ConvoSearch

Jan 2024 – May 2024

Software Developer Intern

  • Integrated GPT-4 with LangChain into an ETL pipeline, hitting 98.3% ingestion accuracy.
  • Built a profit-driven reranking layer that lifted CVR by 18% and order value by 25%.

GPT-4 · LangChain · MongoDB · ETL

Geotab

Summers 2021 & 2022

Junior Summer Program Intern

  • Built Python crash simulation software to model driver safety scenarios.
  • Created a talent dashboard for HR tracking 30+ open roles at once.

Python · Data Visualization · Dashboards

Tools

Languages
Python, Java, C#, TypeScript, JavaScript, Dart, Swift, SQL
Frameworks
Spring Boot, React, Next.js, Node.js, Flask, FastAPI, .NET Core, Flutter
AI / ML
OpenAI API, LangChain, PyTorch, TensorFlow, Hugging Face, ONNX, MCP
Infrastructure
AWS, Azure, Docker, Kubernetes, Kafka, PostgreSQL, MongoDB, CI/CD