Applied AI engineering & consulting

Founder-led / Philadelphia area

AI systems for the work your team still does by hand.

I build practical AI and automation systems for businesses with complex workflows, disconnected software, and expensive manual processes.

For business owners and leaders in operations, finance, and technology.

Where work gets stuck

Important work.
Unnecessarily difficult processes.

Too much work trapped in spreadsheets

Recurring reconciliations, manual reporting, and the same information copied between systems.

Documents that consume staff time

Invoices, contracts, claims, and reports that someone has to read, extract, and check by hand.

Systems that don't communicate

A workflow that needs to read an operational database, call an API, and route exceptions to the right person.

Decisions made with incomplete information

Pricing, cash-flow, or capacity decisions where the data sits in different systems and uncertainty needs to be made explicit.

How I approach the work

The business case comes before the build.

A useful system has to fit your operations, your data, and the consequences of getting something wrong. Sometimes the right answer is simpler software. Sometimes AI makes the difference.

  1. Understand the problem

    Map the actual workflow, find the bottleneck, and establish a baseline. Assess the time, cost, and risk involved before deciding what is worth building.

  2. Build the right system

    Connect databases and APIs, automate handoffs, extract information from documents, or develop agents and analytical tools. Keep people involved where judgment matters.

  3. Measure the results

    Test against real examples and difficult exceptions. Evaluate accuracy, operating costs, and reliability. Set up monitoring so problems remain visible after launch.

Selected work

Systems built. Questions tested.

Independent software, internal development, and research. These examples are not client engagements.

01 / Internal engineering project

Automated Content Intelligence

I built a system that searches TikTok and YouTube for unusually successful content, including outliers from smaller accounts, and monitors X for trending subjects.

It identifies promising opportunities, generates scripts, and integrates ElevenLabs voice technology with short- and long-form video production. Content goes through human review before publication.

Why this matters for operationsThe engineering work is in collecting signals, connecting external tools, and coordinating AI steps with human review. Similar integration problems arise when an agent reads documents, queries business data, and proposes an action that someone needs to approve.

System workflow / Human review before publication
  1. Discover signalsTikTok and YouTube outliers; trending subjects on X
  2. Select topicsIdentify promising content opportunities
  3. Generate scriptsDevelop the selected topics into scripts
  4. Produce voice and videoElevenLabs voice; short- and long-form video
  5. Human reviewReview the content before publication
  6. PublicationOnly after human review

02 / Independent software project

SizeMyGenerator:
Making complex calculations useful.

I built a generator sizing calculator that turns household electrical loads, motor starting requirements, and HVAC equipment inputs into backup power planning estimates.

Homeowners can compare load management scenarios and account for local elevation and temperature using GeoNames and NOAA data. The application brings quantitative modeling and external data into an interactive tool, with assumptions and limits visible alongside the results.

Try the generator sizing calculator

03 / Research

AI Search Visibility Research. Repeated observations across four AI search platforms, with reviewed classifications and explicit limits on what the results can tell us.

Read the research note

A founder-led practice

Engineering judgment.
Quantitative discipline.

William (Bill) Christie FSA, MAAA

Fellow of the Society of Actuaries
Member of the American Academy of Actuaries

BBA in Actuarial Science
Temple University, 2008

Connect on LinkedIn

I'm Bill Christie, an actuary, business operator, and software builder. I have approximately 18 years of actuarial and analytical experience, and currently serve as a Chief Actuary in healthcare insurance and underwriting.

That work spans pricing, financial models, risk assessment, and large datasets. I also spent approximately ten years owning and operating an HVAC business. I question assumptions and measure results because the consequences matter.

Applied AI is a more recent extension of that experience. I build with Python, SQL, APIs, and modern web technologies, combining quantitative judgment with hands-on implementation. At Firefly Clover, you work directly with me.

Firefly Clover is my independent business and is not affiliated with my employer.

Have a process that should work better?

Tell me where the work gets stuck. We'll discuss whether automation or AI can make a meaningful difference.

Discuss a project