rakshit

rakshit k jha

senior software engineer based in london, uk

experience

Meta

Sep 2022 - Present

Senior Software Engineer

  • Currently working in the Account Security Team to increase user security and reduce compromise rates.
  • Led the enablement of Passkeys (biometrics-based auth) for Facebook Web for creation and login.
  • Built heuristics mapping to reduce account compromise from recycled domains and emails, integrating with Google RISC API.
  • Previously improved Appeals Infra and UI to unblock $10-13M/year of annualized incremental revenue.
  • Ensured compliance with EU DSA regulations and worked on regulatory verification requirements in the Authenticity team.
  • Tech Stack: Hack (PHP), Python, Bloks, React, GraphQL, Relay.

FifthRow Technologies

Jun 2021 - Sep 2021

Software Engineer Intern

  • Migrated landing page from Hugo to Next.js and Sass to styled-components, improving Lighthouse performance and SEO to above 90.
  • Developed HTTP REST endpoints for GDT token withdrawals using TypeScript, Node.js, and Express.
  • Developed a monitoring system in Python for liquidity and spread of cryptocurrency perpetuals.

FifthRow Technologies

Jun 2020 - Sep 2020

Software Engineer Intern

  • Automated the colo release process, increasing release rate to three times a week.
  • Designed a Python Client Library for WebSocket API, currently used by over 1000 users.
  • Scripted in kdb+/q and Python to generate reports for market makers.
  • Created a deep learning model deployment pipeline for predicting short-term price movement using Python and TensorFlow Serving.
  • Implemented an OpenAI-based Twitter bot using CoinDesk/Coin Telegraph tweets and a QA bot for affiliates.

Globe Research

Nov 2019 - May 2020

Researcher

  • First author on "Deep Learning for Digital Asset Limit Order Books".
  • Collaborated to forecast short-term price movements of cryptocurrencies using Temporal CNNs.
  • Achieved 0.73 F1-score on walk-forward testing.

education

university of cambridge

oct 2018 - jun 2022

ba & meng in information and computer engineering

Grade: 2.1 (69.92%) (Ranked 86th out of 326)

Key Coursework: Algorithms, Data Structures, Probabilistic ML, Deep Learning, Computer Vision, 3D Graphics.

Master's Project: Gaussian process factor analysis models for neural data (Grade: Class I).

projects

personal blog

[ typescript, next.js, react, redis ]

blogging site with markdown support, serverless rendering, and local caching using redis.

image compression

[ python, pytorch, numpy ]

compressed grayscale images to <5kb using lapped bi-orthogonal transform. voted 2nd highest for visual quality.

agrio

[ django, python, postgresql, react ]

grocery buying app. built backend with django/postgis. voted highest project among 92 entries.

skills

languages

python, rust, sql, bash, c++, c, java, matlab, kdb+/q

web development

javascript, typescript, react, next.js, html, css, tailwind

data & ml

pytorch, tensorflow, numpy, scipy, pandas, xgboost, scikit-learn

infrastructure

git, aws, gcp, docker, mongodb, postgresql

publications

  • r. jha, m. d. paepe, s. holt, j. west, and s. ng. deep learning for digital asset limit order books, 2020.