Implementation
Discovery, requirements, workflow builds, training, and support after go-live. I've done this for five years, usually with several customer projects running at once.
Austin Pierce · Workflow automation and SaaS implementation
I've spent 10+ years across data engineering, product management, and SaaS implementation. Today I run contract lifecycle management (CLM) implementations for customers ranging from mid-market companies to the Fortune 100. I also use AI coding tools (Cursor, Claude Code, Codex) to build internal tools and web and iOS apps.
Discovery, requirements, workflow builds, training, and support after go-live. I've done this for five years, usually with several customer projects running at once.
Contract routing workflows, batch jobs that update tens of thousands of records, and the error handling around them. Before consulting I was a data engineer for three years, writing SQL, stored procedures, and SSIS packages.
I use them to build internal tools and web and iOS apps, including the Python tools I use at work. I don't hand-write production code. I write the plan, review what the agents produce, and test it before it ships.
I lead CLM implementations for mid-market and enterprise customers, up to the Fortune 100.
Product manager on a client project that moved an in-person enrollment process online.
Researched portfolio companies and industries for the investment team.
Data engineer on the team behind the company's operational and reporting data.
B.S., Information Systems Technologies, Minor in Marketing. Southern Illinois University.
Things I've built on my own time.
Cursor implements a written plan, Codex reviews the change, and Claude settles disagreements between them. Nothing merges or deploys until CI passes and the release gates clear, and a release is blocked if an acceptance criterion has no test that ran.
One versioned set of rules that the coding agents in all my repositories follow, covering architecture, database security, design, and error monitoring. A sync script flags repositories that have drifted, and bugs I hit in production turn into new rules.
I've taken several apps from idea to TestFlight and production with React Native, Next.js, and Supabase. They ship with over-the-air updates, analytics, and error monitoring, and one includes a native iOS share extension.
I built a tool that runs the same prompt against Anthropic, OpenAI, and Gemini models and records the output, latency, tokens, and cost of each call. I used it to move one app's feature to a cheaper model, then rewrote that feature's system prompt to resist prompt injection.
One Python pipeline pulled place recommendations out of about 70 saved travel videos, removed duplicates, and put them on a map. Another reads business websites and uses an LLM to save the details to Postgres, switching to a headless browser when a page needs it.
I've run a pool towel brand since 2019. I sampled and sourced from overseas manufacturers, built the Shopify store, set up a fulfillment partner, worked with content creators, and ran the Meta ads. Ad management now runs through an AI process I set up, with Python scripts on the Meta Graph API that pull and audit the ad data.
Lights that follow presence and know whether someone is in bed, plus leak, air-quality, and laundry alerts. I keep the automations as YAML files and push them through the Home Assistant API.
I've been building things on the side since high school.
I built and monetized my first websites at 14, and was sourcing products from overseas manufacturers in high school.
I worked through Ruby on Rails courses at night. I could follow along, but I struggled once I tried to change the code for my own ideas.
I switched to Bubble and launched projects for fun. One was a fantasy football league hub. Another matched Spotify listening history with Ticketmaster and SeatGeek tour dates and texted people through Twilio when an artist they listened to announced a show nearby.
I started a pool towel brand and ran every part of it myself, from sourcing to ads. It's still running, and it's covered in the side projects above.
They let me build the apps and automations I used to only plan, including the tooling that now runs parts of my own business.
From the first discovery call through the support questions after go-live.
If I've done something by hand twice, I look for a way to script it or run it as a batch job.
Nothing an agent writes ships until I've reviewed it and the tests pass.
Before a plan is final, I try to find the case that breaks it.
Select a skill to see where I've used it.
Build CLM workflows in the visual builder, using C# for complex logic, XPath in Word templates, eForms, the platform's APIs, and Salesforce integration.
Built a batch workflow that read a CSV of contract, account, and opportunity IDs and used the CLM's Salesforce ETL step to backfill metadata on 75,000 contracts, with error handling and alerts.
Built Python tools that search workflow definitions and parse configuration packages, which replaced clicking through the workflow designer by hand.
A multi-agent build pipeline. Used in 6 of my repositories.
Home automation. 65 automations.