dhruvdave.ca

Computer Science · Queen's University · Kingston, ON

Dhruv Dave

Interested in the parts of a system users never see.

This past summer I worked at FGF Brands on Cerebro, the platform that monitors production across its 22 bakery plants. The work covered OEE telemetry models, Power BI reporting, and SQL diagnostics against a production database carrying long-standing inconsistencies. Before that: research at the University of Toronto, a web development internship, and a season repairing e-scooters.

Most of the projects below began as questions the documentation did not answer: how a rate limiter stays correct under concurrent load across six APIs, or what a password manager actually does with a user's key. I am in my third year at Queen's and looking for a software engineering internship. If you are hiring, please email me.

Dhruv Dave
Dhruv DaveKingston, ON
Languages
Python, JavaScript, TypeScript, Java, C, SQL
Frameworks
React, FastAPI, Node.js, Hono, Flask, SQLAlchemy, Tailwind CSS, Pandas, NumPy
Databases
PostgreSQL, Redis, SQLite
Tools
Git, Docker, Bash, Power BI, Chrome Extension APIs

Projects

Four of them
01

Markr

Stock screener with a six-source data fallback chain.

Six market-data APIs behind one FastAPI backend. A Redis Lua script enforces each source's rate limit atomically, so a burst of traffic falls through the chain rather than failing on a 429. Candles write asynchronously to monthly-partitioned Postgres with per-resolution Redis TTLs; RSI, MACD, and Bollinger Bands are computed client-side. A Vitest suite covers the fallback chain, the indicators, and the limiter.

  • React
  • FastAPI
  • PostgreSQL
  • Redis
  • Docker
Markr's screener: a search field above a sortable ticker table listing price, change, and market cap, beside watchlist, portfolio, and upcoming-earnings panels
02

SecureVault

Password analyzer and encrypted vault.

Built to understand how a password manager such as 1Password works internally. The analyzer computes entropy, detects keyboard walks and date patterns, and scores passwords from 0 to 100. Each vault entry has its own AES-256-GCM key (PBKDF2, 600,000 iterations), breach checks use HaveIBeenPwned's k-anonymity model so no password leaves the machine, and authentication is bcrypt with rotating JWTs and TOTP 2FA, all covered by a pytest suite.

  • Flask
  • React
  • SQLite
  • Docker
SecureVault's analyzer: a password field that scores entropy and structure as you type, with a numbered sidebar for analyzer, generator, and breach check
03

ctx

Local-first browser session tracker.

page DOM privacy guard password + card scan MV3 extension chrome.alarms batches storage.local works offline Hono API · CLI optional
The page DOM is read through a privacy guard that scans for password and card fields before anything is captured. The guard feeds a Chrome MV3 extension, which batches writes with chrome.alarms into chrome.storage.local so the tracker works offline. A Hono API and CLI sit on an optional second path.

A session tracker that operates entirely offline. The Chrome extension stores everything locally and the server and CLI are optional. A privacy guard scans each page for password and card fields before anything is captured; blocked pages are logged without a URL, title, or content.

  • TypeScript
  • Chrome MV3
  • Hono
  • SQLite
04

Wavvy

Drum machine for the browser.

16-step grid melody pads Tone.js transport clock analyser node Web Audio visualizers 5 modes · canvas PostgreSQL saved patterns
A 16-step sequencer grid and melody pads both feed the Tone.js transport clock, which drives a Web Audio analyser node; five canvas visualizer modes read from the analyser, and patterns are saved to PostgreSQL.

A 16-step sequencer and melody pads scheduled with Tone.js, with five visualizer modes drawn live from the Web Audio analyser node.

  • React
  • TypeScript
  • Tone.js
  • PostgreSQL

Experience

05 roles
  1. May – Aug 2026

    Cerebro AI Technical Specialist FGF Brands

    • Won first place at FGF's 4th annual FGFLife Hackathon with BakeryInsight, a trend-intelligence dashboard that classifies emerging food trends across 9 data sources and scores them on growth, saturation, and regional spread. Each trend is paired with a 6-dimension feasibility assessment (manufacturing, supply chain, cost, safety, brand, shelf life) evaluated against FGF production-line and ingredient data. Built with React 19, TypeScript, Node.js, and Express.
    • Built internal tooling in Node.js and Python documenting Cerebro's OEE telemetry models, changeover logic, and troubleshooting decision trees, along with Power BI dashboards fed by Cerebro's production data across all 22 facilities.
    • Wrote and optimized SQL diagnostic queries against Cerebro's production database to detect and reconcile data anomalies. The queries identified systemic misconfigurations in workflow state transitions and gaps in the upstream pipeline across more than 2,000 production records.
    • Configured and maintained Cerebro environments across FGF's Canadian facilities, and debugged the integration boundary between Cerebro and upstream ERP systems.
    • SQL
    • Node.js
    • Python
    • Power BI
  2. Jul – Aug 2025

    Mobility Hardware and Software Technician Bird

    Inspected, diagnosed, and repaired e-scooters; field-tested new features for the product and engineering teams.

  3. Apr 2025

    Deputy Returning Officer Elections Canada

    Administered a polling station on federal election day.

  4. Jan – Aug 2024

    Research Assistant COBWEB, University of Toronto

    • Implemented agent-based simulations modelling emergent behaviour in ecological adaptive systems with Dr. Brad Bass; configured multi-condition experiments whose results fed the team's research.
    • Wrote Pandas and NumPy pipelines to clean, normalize, and visualize multi-run simulation outputs, revealing population dynamics and threshold-crossing patterns across conditions.
    • Python
    • Pandas
    • NumPy
  5. Sep 2022 – Feb 2023

    Software Engineer Intern Digitera

    Developed and shipped web applications for small-business clients using JavaScript, HTML, and CSS; improved page performance by 40% through asset minification, image compression, and deferred script loading.

Education

Sep 2024 – Present
  1. Sep 2024 – Present

    Queen's University Kingston, ON

    Bachelor of Computer Science (Honours), Mathematics minor. GPA 3.8.

    • Data Structures & Algorithms
    • Computer Architecture
    • Discrete Structures
    • Linear Algebra
    • Differential Equations

Contact

open to internships
davedhruv2006@gmail.com

Email is the best way to reach me. I read every message. Away from a keyboard, I am usually playing hockey, out on a trail, or travelling.