What We've Built
Real systems. One goal: more output.
These are real systems running in production today; not concepts. Here's what they actually do.
News Publishing Case Study
A newsroom that never misses a local story; and never sleeps.
SignalsAZ.com draws from municipal and chamber news sources across the region, continuously. Rather than someone manually checking dozens of sources every day, an automated system monitors them, brings qualified stories into our editorial workflow, and prepares a fully formatted, ready-to-review WordPress draft; matched with a relevant featured image chosen from our own licensed photo library, not a generic stock photo.
The result: a steady stream of local coverage that's ready for an editor's eye, without anyone spending their morning hunting for what happened overnight.
Local sources monitored→
Story qualified→
Draft written & formatted→
Ready for editorial review
Podcast Production Case Study
From raw transcript to fully packaged episode; automatically.
Every CAST11.com episode used to mean a manual slog through the same steps: writing show notes, drafting an article, pulling social copy, prepping SEO metadata, cutting clips. We replaced that with a single automated pipeline that picks up the moment a transcript is uploaded and produces the full package on its own; over 20 separate publishing assets per episode, generated and organized without anyone touching a keyboard.
A companion tool we built in-house, nicknamed "Demux," handles the front end of the process: it pulls the audio and transcript straight out of the raw episode video, quietly uploads the episode to YouTube, and kicks the transcript into the pipeline; so the moment recording wraps, production has already started.
Transcript uploaded→
AI drafts 20+ assets→
Episode tracked automatically→
Video clips queued
Website Development Case Study
A full website redesign – built hand-in-hand with AI.
The site you're on right now was redesigned in-house, with AI as a hands-on development partner. Working with Claude, we planned, wrote, and refined every page template ourselves: a custom WordPress theme with its own dark, cinematic design system, built and tested on a separate staging copy of the site before anything touched the live one. Every change went through that same loop.
It was built to stay easy to run. Portfolio pieces, team members, and blog posts all pull straight from WordPress, so anyone on the team can add or update them from the dashboard without touching code. A single shared link file keeps navigation consistent across every page, and heavy media was optimized along the way; one studio video clip went from 70MB down to 3.2MB so pages load fast.
Design planned→
Templates built with AI→
Tested on staging→
Reviewed & refined
App Development Case Study
A grant-tracking app for a regional coalition – built with AI.
GRyT, a Prescott-area regional coalition, received a USDA Rural Business Development Grant to carry out a 12-month regional plan, and needed a way to keep its board and partners accountable to it. Their starting point was a single-file prototype that only worked on one computer, with no logins. Working with Claude, we turned it into a full multi-user web app, built in Lovable on a Supabase backend, so everyone works from the same live data.
The app tracks the grant end to end: a dashboard covering nine KPI areas with green, yellow, and red status, an action register built around the rule that every item has a named owner and a due date, a budget tracked against the USDA-approved categories, a 90-day launch plan, and the coalition's five working groups. Admins manage who has access, and everything can be exported or backed up at any time.
Grant plan mapped→
Built with Claude + Lovable→
Multi-user logins→
Refined with coalition leaders
Company-Wide AI Rollout Case Study
140+ AI workflows across our parent company – one at a time.
We're now helping our parent company, Fain Signature Group, implement more than 140 AI workflows across the business, from lease renewals to client acquisition. Each one is triggered automatically – by a new document, an email, or a schedule – then analyzed by Claude and delivered as a ready-to-use summary card in that team's private Microsoft Teams channel.
We're just getting started. The first workflows automate a lease-tracking app that one of FSG's own team members built to keep tabs on client leases. Every workflow is tested in real use by the people who rely on it, refined from their feedback, and only goes live once it's approved.
Trigger fires→
Claude analyzes→
Card posted to Teams→
Tested & approved by the team
