Senior Software Engineer on Data Infrastructure (RDBMS) at TRM Labs

TRM Labs provides AI-powered intelligence solutions that help public and private sector agencies investigate and disrupt crime. TRM's platforms enable investigators to trace illicit activity, build cases, and construct operating pictures of threat networks. Leading agencies and businesses worldwide rely on TRM to make the world safer and more secure.

The Data Platform team builds and owns highly available, scalable data infrastructure for TRM's products and services. As a Senior Software Engineer on Data Infrastructure (RDBMS), you will develop, operate, and scale the relational database systems that serve data at petabyte scale, helping build a safer financial system for billions of people. This role is uniquely broad at TRM: you own the serving layer end to end, from query performance to cost to availability, on infrastructure that product teams depend on every day.

Key Impact Areas:

• Keep customer-facing APIs fast and available at five-nines by owning query tuning, index design, and schema optimization on petabyte-scale Postgres/Citus, directly determining whether product teams can serve data in real time.

• Cut storage and compute costs materially by using AI-assisted analysis (pganalyze paired with Claude) to detect compression and data-model opportunities, then shipping validated changes to production.

• Remove team toil by building agentic automation for routine database operations, such as self-serve pgbouncer provisioning, disk scaling, and blue-green deployments, so any engineer can run them safely.

• Protect data integrity and latency by driving agentic validation of database changes before they reach production.

• Keep real-time data flowing by managing CDC pipelines (PeerDB, Fivetran, Debezium) and using AI tooling to debug replication failures faster.

• Shape the next-generation platform by migrating workloads off first-gen infrastructure and prototyping new data stores with AI-accelerated spikes.

• Raise the whole team's leverage by codifying your workflows into AI-native runbooks and internal tooling that teammates reuse.

Required Qualifications:

• 5–8 years building and operating production PostgreSQL (Citus, Aurora, AlloyDB, or equivalent distributed Postgres).

• Deep SQL optimization skills (Explain Plans, CTEs, window functions, partitioning, index design, query-planner behavior in distributed environments), increasingly paired with AI-assisted query analysis.

• Hands-on experience with CDC tools (PeerDB, Fivetran, Debezium, Datastream, Airbyte) and comfort using AI tooling to debug replication failure modes.

• Fluency with database profiling (pganalyze or equivalent) to interpret metrics and logs, including using LLMs to summarize performance findings.

• Production automation experience in Python or Go, including agentic automation of routine database tasks; Postgres extension development is a plus.

• Daily use of AI coding tools (Claude, Copilot) to accelerate development and produce higher-quality output faster, with the judgment to know when to trust or reject AI output.

• A track record of using AI-assisted analysis to drive measurable cost or performance improvements in production database systems.

About the Team:

• The Data Platform team is distributed, not distant. We operate async-first across US and Canada working hours with intentional synchronous touchpoints when speed demands it.

• We own the petabyte-scale data infrastructure that powers TRM's products. The scope is large, the mission is real, and the work is technically demanding.

• Team culture is product-focused, customer-focused, coaching-oriented, scrappy when needed, and process-oriented when executing against OKRs.

• You will work alongside senior engineers who care deeply about their craft and hold high standards for code quality, operational excellence, and cross-team communication.

Team Operating Rhythms:

• Distributed, not distant: most collaboration happens async in Slack, Notion, and GitHub, and meetings are purposeful and minimal.

• Primary timezone overlap across US and Canada working hours; our north star is no IC spending more than 3–4 hours per week in meetings.

• Written-first culture: technical specs and decision docs drive alignment before code.

On Call Requirements:

• On-call rotation (approximately once every 15 days; each shift is a 24-hour window across 2 days).

Expected Impact Within 90 Days:

• Eliminate a single point of failure by shipping self-serve scaling automation (pgbouncer provisioning, disk resizing) that any teammate can run.

• Identify a compression opportunity with AI-assisted analysis, validate it with a proof of concept, and ship it to production for a measurable storage-cost reduction.

• Coordinate a schema change across three product-engineering teams in parallel and deliver the platform changes ahead of a product deadline.