Data Scientist, Data Quality and Benchmarking
About the Role
You will build and operate real-time pricing systems that produce high-integrity reference prices across thousands of assets. You will design models and detection logic to keep prices trustworthy under stress, model adversarial market behavior to locate risk, and lead data forensics during incidents (pre/post-trade analysis, benchmark vs pool deltas, liquidity gaps). You will quantify user impact and implement fixes that measurably reduce recurrence, and contribute to alerting and monitoring for data quality and benchmark health.
Requirements
- Working knowledge of DeFi primitives
- Oracle literacy including CEX/DEX sourcing and median/TWAP construction
- On-chain data fluency including reading EVM logs and traces and using ABI and token standards
- Ability to identify swaps transfers and liquidations in event streams
- Familiarity with subgraph and indexing patterns
- Understanding of block production finality and reorgs, mempool dynamics, and gas markets
- Market microstructure knowledge including liquidity slippage pool imbalance funding rates and open interest
- Experience reading simple smart contracts (Solidity or Rust)
- Experience with RPC providers and indexing stacks such as subgraphs and BigQuery
- Experience designing anomaly detection for financial time series and SLO-driven alerting
- Experience with monitoring and on-call tools such as Prometheus Alertmanager Grafana and PagerDuty
- Proficiency in Python and SQL and experience with web3 libraries and ABI decoding
- Experience with Dune Flipside Covalent or BigQuery for blockchain analytics
- Experience with Kafka or other stream processing systems
- Portfolio showing dashboards alerts price benchmarks and data-quality detectors
Responsibilities
- Build real-time pricing systems
- Design models and detection logic to maintain price trustworthiness
- Model complex adversarial markets to identify risk
- Lead data forensics and incident investigations
- Quantify user impact and implement fixes to reduce recurrence
Skills
Pattern RecognitionErc-20FlinkMempoolPagerdutyBlock-ProductionSlippageErc-721Open InterestCovalentFlipsideAbiOracle PricingFinalityStream ProcessingErc-1155Chain Re-Org DetectionEvmRpcIndexingBigqueryDuneIncident ResponseSolidityOracleSubgraphAlertmanagerSloRustDefiOn-Chain DataPost-Trade AnalysisGrafanaPrometheusPerpetualsTime SeriesSqlPythonLiquidityReorgsGas MarketFunding RatesAnomaly DetectionPrice BenchmarkingData ForensicsKafkaWeb3