Data Engineer, Indexing and Market Data

About the role

Alpha’s market data is first party. A Rust indexer follows Solana over Yellowstone gRPC and decodes swaps across Pump.fun, Raydium, Meteora, Orca and the launchpads into a swap ledger past a billion rows, candles from one second to one week, and per-wallet positions. Sibling pipelines cover EVM chains and TON. On top sit the price plane, cost basis and P&L, spam and scam filtering, and the known-wallet tape that powers Radar and copy trading.

This role owns the correctness and latency of that stack. A candle that is off by a fee, a position that double counts a decimal change, a price that is an hour stale: those are the bugs you will hunt and design away.

What you will do

• Decode new programs and venues as they launch, and keep the decoders pinned to fixtures so upstream changes cannot silently change our numbers.
• Keep ingest at slot level: streaming from gRPC into synchronous hot paths, dedup on the way in, back-pressure that never blocks the trade path.
• Own candle, position and cost-basis correctness, with parity checks against independent sources and a written incident ledger.
• Build wallet intelligence: cohort scoring, forward-return measurement, sandwich and bot filtering, the signals that tell a user whose trades are worth following.
• Scale the store: TimescaleDB hypertables, retention, compression, and query shapes that stay fast at ten times today’s volume.
• Publish the data to the product: gRPC services, snapshots, Redis fan-out, and the docs that explain what each number means.

What we are looking for

• Production experience in Rust or another systems language, with real streaming or high-throughput ingest behind you.
• SQL at scale: Postgres or Timescale, query planning, partitioning, and the discipline to measure before optimising.
• Working knowledge of how DEXs actually work: constant-product and concentrated liquidity, bonding curves, routing, fees, MEV.
• A correctness reflex. You reconcile against a second source before you trust the first.
• Comfort turning a vague product question into a dataset, a metric and an answer.

Nice to have

• Solana program decoding (Anchor IDLs, inner instructions, Geyser or Yellowstone plugins).
• EVM log and trace ingestion, or TON message parsing.
• Quant or research work: cohort analysis, backtests, forward returns.
• Python for analysis alongside Rust for production.
• Kafka, Debezium or other CDC pipelines.

We encourage you to apply even if you do not meet every line above. Strong people rarely tick every box.

How to apply

Email us at hi@tryalpha.xyz with “Data Engineer, Indexing and Market Data” in the subject line, a link to work you are proud of, and a few lines on the hardest system you have shipped and what broke. No cover letter, no form. A real person reads every application.

Apply by email