Chenxiang Li
Experienced crypto researcher and data analyst at top CEX and DEX. Skilled in SQL, Python, subgraph queries, and Google Analytics. Experienced in driving business impact with data insights. Ability to work remotely with cross-functional teams in English.
Experience
Data Analyst
Apex Exchange
- Liquidity Data: analyzed CLOB liquidity and market quality, including spread, depth, slippage, top-of-book stability, and no-trade gaps; worked with market makers to diagnose liquidity issues and improve symbol-level order book performance
- Data Agent: built an internal Lark-based data agent that converts business users’ plain-language analytics questions into SQL via Codex and returns data outputs from Redshift with insights
Individual Advisor & Researcher
Individual Advisor & Researcher
- Mantle Ecosystem Overview: to track ecosystem level performance and key sectors
- AAVE Liquidations & Staking: to analyze liquidations and staking products of AAVE
- Monad Testnet Ecosystem Top Projects: to track performances of top projects on Monad testnet
- Onchain Stock Token Metrics: to track the onchain performance of xStocks tokens on Solana and EVM chains
- Magic Eden $ME Airdrop on Solana: to decode Solana on-chain event logs to visualize on-chain distribution metrics, user engagement trends, and token claim patterns
Co-founder
Co-founder of a cross-chain DEX on Arbitrum
- Focused on data and product and GTM strategies reaching TVL to $4m
- Fostered Partnerships on cross-chain capabilities (LayerZero, Symbiosis), aggregator integration (Odos, OKX DEX), liquidity bootstrap (Frax and meme coins), analytics (DEX Screener, GeckoTerminal)
- Industry Engagement: Participated in Web3 conferences (TOKEN2049, DuneCon, DevCon, etc.) and engaged with developer communities, stay updated on industry trends, and identify potential collaborations in the DeFi and web3 data analytics
Data Scientist - PancakeSwap
Binance
Remote
- Built data pipelines with SQL , Python, and subgraph queries from on-chain data providers (Dune API, GraphQL, Bitquery, etc)
- Designed and developed CAKE Tokenomics Dashboard and BNB Chain DEX Dashboard on Dune Analytics
- Made data-driven recommendations to identify early-stage projects on BNB chain and Aptos to for BD team across IFOs, Farms and Syrup Pool products
- Led a team of engineers and collaborated with Chainlink integrating Automation and price feed to build Prediction and Lottery products in a decentralized and cost-efficient manner
- Proven project management experience in coordinating a cross-functional team, prioritizing tasks and monitoring process to ensure on-time delivery of CAKE buybacks, Monday Burn
Business Data Analyst - Central Marketing
Binance
- Designed and organized Trading Competition and Token Task Campaign for user acquisitions of listed projects
- Designed and implemented A/B tests to gain insights into user conversion: Signup -> 2FA -> KYC -> Deposit -> Trade
- Collaborated with cross-functional teams to identify potential user growth opportunities, track campaign performance
Senior Data Analyst
Kwai (TikTok competitor)
Beijing, China
- Designed and evaluated A/B tests for new product features regarding share, download and social connections
- Design and develop metric system to evaluate and monitor the quality and quantity of social connections
Senior BI Algorithm Engineer
Danke Apartment
Beijing, China
- Renting room data diagnosis system: Root-cause analysis based on general renting markets, price monitoring and business data
- Prediction of turnover probability: Apply machine learning models to predict potential unsalable rooms
Business Operations & Strategy Associate - Global Business
ofo
Beijing, China
- Smartmap - First Geohash-based Decision Support Map Tool ■ Designed and developed prototype, based on Geohash grid to visualized operations metrics with color, label and shape ■ Marketed and trained local operations team for how to use Smartmap to plan their operations work
- SCAN based Data-driven Operations System ■ Supply side: Based on Smartmap and BI system, quantify business performance and efficiencies from spatial and temporal ■ Ops Team: Built SCAN based operations data system, got involved in design and development of internal operations tools
Data Analyst - Strategic Operations
ofo
- Demand side: ■ Apply Growth Hacking techniques and machine learning models to find potential user growth opportunity ■ Design and analyze A/B tests to drive product iterations and design user operation strategies ■ Apply AARRR, RFM, etc. methodologies to drive user conversion and retention
- Supply side: ■ Developed subway station based operations data system to derive insights and drive decision-making for operations
Data Analyst
Gravity4 (An online advertising startup)
San Francisco, CA
- Extracted data by complex SQL queries from multiple databases, aggregated and manipulated using Python Pandas
- Designed A/B tests for new technical features, extract data with ad-hoc SQL queries for reporting and QA
- Productized statistical results in Tableau to monitor traffic metrics and make statistical analysis repeatable to all campaigns
Education
Arizona State University
Master of Science
GPA: 3.7/4.0
Course: Regression Analysis; Data Mining; Information System; Operation Research; Design of Experiment; Financial Engineering; Enterprise Modeling; Stochastic Simulation; Quality Control
Nanjing University of Finance and Economics
Bachelor of Engineering
GPA: 3.4/4.0
Excellent Learning Scholarship, Third Prize (Top 25% GPA)
Skills
PROJECT EXPERIENCE
ofo - Smartmap (Geohash-based Data Visualization Tools on Google Maps
Worked with data engineering team to develop UDF for HIVE to aggregate location-based data by Geohash
Developed and designed prototype with Google Maps API to visualize business metrics with label, color and shape
Communicated and trained with city operations team to iterate product according to their feedback
ofo - Predicting ofo Rider Churn + Intelligent SMS Loyalty Program
Feature Selection: Constructed 20+ user features of user personas and RFM(Recency, Frequency, Monetary)
Data Preprocess: Filled missing values, removed outliers, encoded non-numerical variables
Modeling: Compared predictive powers of Logistic Regression, Random Forest and GBDT classifiers by sklearn
Parameter Tuning: Tuned parameters (n_estimators, max_features, max_depth) of Random Forest by Grid Search
Model Evaluation: Evaluated models by AUC scorer and cross-validation, finally compared feature importance
Action Plan: Designed A/B tests to evaluate model reliability and different means of user loyalty
ofo - Find User Retention Magic Number by Logistic Regression
Trained a LR model in Python on new user behavioral data (RFM) to investigate user action and retention
Checked coefficients for important user actions and verified with backward A/B tests