I'm Asia Stewart. I help startups and small-to-medium businesses turn AI from a buzzword into working systems: agentic workflows and tools tied to revenue, lower cost, and added capacity. Ten-plus years building businesses and the operating systems behind them, including five years as a platform Senior PM at American Express.
The firm wanted to test a new service line but had no offer, no tooling, and no efficient way to fill the funnel. I developed the offer itself, built the client-facing tool behind it, and deployed a system of agents to source clients, qualify leads for fit, and automate document prep so attorneys review and finalize only the last 20 percent.
I also built an ad-optimization workflow that lifted lead generation while cutting ad spend nearly in half, making the new vertical viable at a flat rate.
ChurchSpace converts underutilized church real estate into micro-fulfillment centers in partnership with last-mile carriers. As the founders' operational right hand, I built the Notion-based operating system, the dashboards four different audiences actually use, and the cadences that stopped leadership from being the bottleneck.
I designed an agent-assisted operations layer, including a church-prospecting agent and a pipeline-monitoring agent that surfaces stalls and drafts weekly updates, so operational strain becomes visible early and continuously.
The platform negotiates medical bills on behalf of patients. After 20-plus discovery interviews with solo, multi-site, and hospital billing professionals, I reframed the positioning from "discount engine" to "payment-resolution tool," changing how the product earned trust.
I ran structured evals on the AI's negotiation recommendations, stress-testing for dollar-amount accuracy, hallucinated CPT codes, and unsafe legal language before any output reached a patient, and built the inference unit-economics model that located the fee threshold where margin breaks.
Every build maps to a number the business cares about: new revenue, lower cost, added capacity. No AI for its own sake.
I work from the offer down through the tool and the agents behind it, so the strategy and the system that runs it stay in sync.
Structured evals and human review gates keep agent output accurate and trustworthy before it ever reaches a customer.