I’m Nathan Slaughter, a software engineer with more than a decade of experience. My work spans fintech, observability, data pipelines, and global infrastructure.
Before software became my main work, I managed investment portfolios. That background taught me to ask what the evidence supports.
Selected experience
- At Lightstep, I managed contractors building metrics integrations across infrastructure including Kafka, Cassandra, and RabbitMQ. The work included two OpenTelemetry Collector receivers.
- I built and operated a metrics pipeline delivering more than 50 MB/s into a data lake.
- I operated the observability components of a Kubernetes platform spanning more than 50 datacenters and helped workload teams deploy, instrument, and run their services.
Read more about my work and background →
Independent projects
My independent practice serves investment managers and trading firms, and the companies that supply their data. For managers and trading firms, I set up Claude with shared skills and read-only access to their data, so the whole office works the same way. I build research tools: backtests, monitoring for live strategies, and alerts on the measures they watch. I also build integrations that bring third-party data into their applications, and I check that the data feeds they pay for arrive on time and complete. For data providers, I run the same checks with a customer’s access. I take on fixed-scope projects, with ongoing maintenance and monitoring available afterward.
Writing
- Designing a financial data API customers can keep in sync
An export has to carry the update cursor that matches its snapshot, or a customer who follows the documentation can still miss a revision.
- Designing a data SDK for downloads that fail halfway
When a download fails on page four, the customer's job needs to know what it saved and where to resume. How to design an SDK's interface, errors, and retries around that moment.
- Checking whether your data feeds deliver what you pay for
A feed can arrive on time and still hold yesterday's data. Check each delivery when it's due against an expectation from outside the feed, and keep what arrived.
- The timestamps that make financial data usable
A backtest needs each number as it was known when the decision was made. Publication dates answer that for nightly strategies; intraday strategies also need the provider's delivery time.
Get in touch
I’m also open to engineering roles where I can help teams build and operate reliable software, across industries.
Contact me on LinkedIn with a note about your team, the role or project, and the problem you’re working on.