me

Boris Topalov

full-stack software engineer & hobbyist musician

new york, ny

[email protected]

about me

I'm a full-stack software engineer based in NYC with 2.5 years of professional experience. I currently work on B2B payments infrastructure at Capital One. Outside of work, I'm learning about LLMs by building Abby, an AI assistant for Ableton, and Petri, a tool for multi-agent AI-to-AI experiments. Beyond coding, I spend most of my time making music; these days I'm mainly producing electronic music and DJing.

work experience

Software Engineer | Capital One
August 2023-Present
  • Automated a previously-manual payment refund process which reduced refund processing time from several days to less than 15 minutes.
  • Built a visualization tool for customer payment lifecycles, which is used by support agents servicing thousands of enterprise customers.
  • Upgraded legacy Java codebases and resolved JVM issues such as memory misallocation inside Docker containers.
  • Created a Quickbooks integration which received an award at 2023 company hackathon.
Software Engineer | Audialab
June 2024-October 2024
  • Owned, designed, developed, and shipped a cross-platform desktop audio plugin built in C++ (JUCE) and React.
  • Built a pipeline in Python and AWS that used machine learning to extract musical stems for over 50,000 audio samples.
  • Rearchitected core state management of a React codebase, resolving long-standing bugs and performance issues.
  • Developed product strategy and pitch decks, created marketing material, and provided direct customer support for our applications on Discord.
Software Engineer Intern | Comm
September 2021-December 2022
  • Built a fully autonomous GitHub crawler & scraper system which parsed thousands of GitHub pages and algorithmically ranked tens of thousands of GitHub users as candidates for recruiting.
  • Created a custom priority-based job queuing system used by the system to prioritize scraping highly relevant data based on user-provided metrics.
Business Analyst | Capital One
August 2022-August 2023
  • Improved the performance of core SQL scripts by ~70% through SQL profilers, and re-wrote large chunks of undocumented, inefficient code into well-documented and maintainable solutions.
  • Created an automation-focused analytics framework which was used by teams covering $20B lending portfolios.
  • Developed a monitoring and analytics dashboard that analyzed millions of data points every day to aid executive decision-making of a $7B lending business.

projects

abby

An LLM music production assistant for Ableton that can create new device chains and make suggestions as you produce.

petri

A framework for orchestrating multi-agent experiments in a user-controlled experiment setting.

locallm

Fully local chat app for an LLM agent that searches the web, cites its sources, and stores information in a vector database (RAG).

wordle-versus

Multiplayer Wordle!

spotseek

Golang client for Soulseek, a distributed peer-to-peer file sharing protocol.

technologies

backend
Go
Python
Node.JS
Java
frontend
React
Svelte
TypeScript
Tailwind
tools + infra
AWS (Certified)
Docker
GH Actions
PostgreSQL
MongoDB

education

University of Virginia B.S. in Commerce, Minor in Computer Science
2018–2022