Jarvis Price
Hi, I’m Jarvis. I automate the workflows a business is still doing by hand, build the dashboards that report on them, and take an MVP from an idea to something running in production.
Get in touchAn MVP lands in days to weeks, depending on scale.
I can take a system from described to running in about the time it usually takes to finish scoping one.
Shipping fast only counts if the thing still works next month. The same build gets retries, error handling, logging and alerts, because an API will change, a token will expire and a job will quietly stop.
Updated August 2026
A website scanning and reporting platform, designed, built and run alone from the engine to the deployed site. A fleet of headless browsers over more than 1,300 live sites, a Fastify API over Redis-backed queues, structured results in Postgres, and reports produced on a schedule with nobody supervising them.
risksignal.io ↗
Lead response for businesses that pay for leads and then lose them to whoever calls back first. Every lead is answered inside sixty seconds and chased, with nobody watching a phone.
The reporting above is a Cloudflare Worker over its own Supabase project: the ingest path, the read API, and the page that renders it. Built as code rather than assembled out of a reporting tool.
The figures above are a demonstration account.
jarvisprice.com/services ↗Takes a list of targets and returns structured, de-duplicated records. It renders each page, uses a language model to pull fields out of what a browser actually sees, resolves duplicates, and batches the result for whatever runs next. Staged, so a run picks up where it stopped.
risksignal.io, the Sixty service pages, and this one. Three separate visual identities on Cloudflare Pages, custom HTML and CSS, responsive down to a 390px phone, with a Worker handling the contact path behind bot verification, per-IP rate limiting and transactional email. No page builder and no template.
Every project carries its own automated review and a set of checks that block a change breaking a rule the project has already set, with the database and hosting wired in directly. A regression gets caught at the point it is written, before it reaches production.
Shipped on TypeScript and Node, Postgres on Supabase with the migrations, Playwright for anything that needs a real browser, BullMQ and Redis when work has to queue and survive a restart, Fastify for the API, and Cloudflare Workers and Pages for everything deployed. Python and Bash for the glue.
Beyond that the stack is rarely the constraint. Next.js and React on Vercel, REST and GraphQL APIs, webhooks and third-party integrations, Docker, AWS and GCP, and whatever automation platform a client already runs. If a project needs something not on that list, picking it up is part of the work.
One of these is a product rather than a project. Businesses pay for leads and then lose them to whoever calls back first, so I built Sixty: it answers every lead inside sixty seconds, chases it, and reports what it produced against what was spent.
See how it works, or read the same system pointed at commercial insurance, where the first quote usually writes the policy.
RiskSignal is the other one that runs: a fleet of headless browsers checking conversion tracking across a marketing agency’s client sites every day, and reporting the day one breaks.
Tell me what you want built, or what your team still does by hand. I will come back with how it would work and how fast it can ship.
Source is private. I am happy to walk through the architecture and the decisions behind it on a call.
contact@jarvisprice.com