work
what sharpening looks like in practice.
A few engagements, laid out as a connected path. Client names and numbers are held back until each study is cleared to publish.
- ai systemsB2B software2025
A support assistant that drafts replies from the docs it can cite.
[METRIC] of tickets now start from a cited draft
A support team was drowning in repeat questions that the documentation already answered. We built an assistant that drafts replies grounded in their own content, and only when it can point to a source.
TypeScriptCloudflare Workersa vector storean LLM provider - automationfintech2024
Nightly reconciliation that replaced a spreadsheet and a prayer.
nightly close cut from [METRIC] hours to zero
A finance team closed each day by hand-matching payments across three systems in a spreadsheet. It was slow, and the errors were expensive. We turned the whole thing into a job that runs itself and flags only what needs a human.
TypeScripta queuePostgreSQLscheduled workers - backend engineeringmarketplace2024
Cutting a core API’s p95 without a rewrite.
p95 latency down [METRIC] with no rewrite
A marketplace’s core API had grown slow and unpredictable as the data grew. Rather than the rewrite everyone dreaded, we profiled it, fixed the paths that mattered, and left it faster and easier to reason about.
GoPostgreSQLRedisOpenTelemetry
systems sharpened
years in the workshop
average latency cut
uptime held under load
in their words
what clients say.
[TESTIMONIAL] - a two or three sentence quote about the engagement in the client’s own words. Replace once approved for publication.
[TESTIMONIAL] - what changed for the team after the work shipped, said plainly. Kept as a placeholder until we have a real, cleared quote.
start here
have a system that used to feel fast?
Tell us what it does and where it hurts. We’ll tell you honestly whether it needs sharpening or rebuilding - and what that would take.