• Terminal + LLM: from Warp to Ghostty

  How I replace Warp with Ghostty.

• Apr2022 - Dec2024
  Freelance
  Crypto Analytics Engineering
  • Helped Oku.trade: research smart contracts, identify onchain users of competitors, build custom dbt data models and dashboard to collect and monitor the users for a user acquisition campaign.
  • Developed and optimized DBT data models for Uniswap, Odos, Jumper (Li.Fi), and Bungee (Socket), enhancing processing efficiency and integration as open source contributor at Dune.com.
  • Uniswap Analytics Grants — 1st Prize Winner (Jun-Aug 2023): Won 1st prize with 2 comprehensive on-chain research analyses:
    • Uniswap Memecoin Mania Analysis: Analyzed PEPE-led meme coin trading surge (Apr-Jun 2023) covering $31B aggregate volume across 4,000+ minor pools. Found meme pools peaked at 21% of daily trading volume and 75% of new traders at height of mania. Identified 180K+ new DeFi users onboarded to Ethereum via meme trading. Analyzed whale behavior (0.2% of traders accounted for 52% of volume), trader profitability patterns (experience beats activeness), and cross-pool liquidity migration effects. Discovered meme traders showed 1.5x higher retention vs average traders in first 3 weeks.
    • Uniswap User Origin Analysis: Deep-dive into trader acquisition routes analyzing 11.3M traders, $1.6T volume, 276M+ trades across 450K+ contracts. Segmented users by trading route: Uniswap Frontend (55% of onboarding, 29% of volume), Aggregators (35% onboarding, 15% volume), MEV (1% onboarding, 42% volume), and Unknown contracts (9% onboarding, 14% volume). Analyzed multi-route trader behavior, cohort retention patterns, and weekly growth rates by route. Built live Dune dashboard for real-time user origination tracking.
• Jun2020 - Mar2022
  TIKI Corporation
  Growth Data Specialist / Senior Growth Analyst

  Led a team of 3 analysts, using BigQuery, Looker, Google Sheets, Airflow, Superset, and Amplitude to monitor user growth.

  • Identified optimal threshold and content with A/B test to maximize notifications open rate, minimize spam (email, app/in-app push).
  • Analyzed journey of first-purchase users to find visits threshold for optimal activation.
  • Analyzed patterns of engaged users to develop strategies for retention and increased lifetime value.
  • Built and managed data model versions with Bitbucket.

• Credit Card Fraud Detection: Rules vs Machine Learning

  Comparing rule-based detection with Random Forest ML on 1.3M credit card transactions
• NYC Taxi Spatiotemporal Analysis

  A comprehensive spatiotemporal data analysis project demonstrating EDA, zone grouping, and behavior-based anomaly detection on NYC Yellow Taxi trip data with interactive Streamlit dashboard.

Let's Connect

• x (twitter) /
• github /
• linkedin /
• quangphu.le512@gmail.com