What Howard is for

Howard is your school business expert. From the day you sign up, it should be one of the most useful members of your team, because it takes on the hardest questions in running a school: the ones caught between federal rules, state law, TEA guidance, what program staff tell you, and your own district's local policy.

People outside school finance tend to assume every question has a black-and-white answer. Many don't. Howard's job is to help you turn those gray areas into decisions you can stand behind, because you know exactly which laws and policies they rest on.

Today Howard covers school finance, accounting, grant management, and compliance. That's the start. Every district, campus, administrator, leader, and teacher works on top of tens of thousands of pages: laws, TEA policy, guidance and trainings, local policies, programs and initiatives, curriculum, and best practices. The long-term goal is for Howard to read and understand all of it, down to the last word on the last page, so that whenever you're working out the best way forward, Howard can help you find the answer.

Who built this?

I built Howard. My name is Donnie Clapp, and I am a Grant/Program/Project Manager at Region 3 Education Service Center in Victoria, Texas. After years in school finance, I got tired of watching how much useful knowledge lives only in the heads of people who are either retiring or too busy to pick up the phone, leaving everyone else to dig through PDFs. I built Howard because finding a straight answer in FASRG or EDGAR should not feel like an archaeological dig, especially for a new business manager.

I named it Howard because Howard sounds like someone who has been balancing ledgers since before you were born. It is also my middle name, and it contains the word how, which is usually what you are asking when you open the thing.

What does Howard know?

• FASRG v20 — Texas Financial Accountability System Resource Guide (v19 kept for historical questions)
• EDGAR & 2 CFR 200 — federal grant regulations and the Uniform Guidance
• TEA General & Fiscal Guidelines — 2025 edition, plus 2021 for history
• LASO Cycles 3 & 4 — program guidelines, FAQs, and errata for every initiative
• State funding — FSP overview, Summary of Finances guide, HB 2 funding changes, budgeting handbook, FIRST
• EMAT — instructional materials certification, IMTA funds, disbursements
• ASCENDER business documentation — the complete business-office wiki: payroll, finance, budget, purchasing, personnel, asset management, warehouse, and more (4,000+ pages)

If the answer isn't in those documents, Howard says so rather than guessing. Know a document Howard should read? Suggest a source.

How Howard keeps its answers honest

I am not going to tell you the AI is magic, and I am not going to pretend it is perfect. Since March 2026, Howard has been rebuilt around one rule: every answer has to point to the page it came from.

• It reads the real documents. Howard reads from a library of 108 official sources, which we have cut up into about 12,500 searchable passages. When it answers, it quotes those documents and links you directly to the specific page on the TEA, federal, or ASCENDER site, so you can check its work.
• A second pass checks the first. But pointing to a page is not the same as being right. Before you see an answer, a separate checker lists every fact it contains and tries to find the source sentence behind each one. If a fact has no support, Howard digs through its library again, then searches a short list of official websites, and either rewrites the answer with what it found or tells you plainly it could not confirm it.
• It knows your words, not only TEA's. We also keep a growing glossary, because people in business offices rarely talk the way the official documents do. If you ask about the comparability thing, or about paying money back to TEA, Howard maps it to the official term (for that last one, Liability object code 2181) and finds the right page anyway.
• It gets tested against real questions. Before a change to Howard's search or to the AI models behind it goes live, we score it against a set of 120 real questions with known right answers, so we can tell whether the change actually helped or just sounded nice.
• Mistakes get fixed, not just noted. The "this answer is wrong" button is not decoration. We look at every single report in a daily review. We check the answer against the source, fix it, re-ask the question to make sure the fix sticks, and add it to the test set so the same mistake does not quietly come back. When TEA's own document has a misprint, we report it to TEA.
• It keeps up with TEA. When TEA reissues a document, or sends a To the Administrator Addressed letter that changes a rule or deadline, we put the new version in Howard's library and retire the old one, so Howard answers from the current document, not the one it replaced.

Who runs it?

Howard is built and maintained by 👏 Applause Consulting, which is my small consulting practice. At Applause, I help small organizations with marketing and strategy, build websites, and write practical software and automation. Howard is what happens when that skillset meets a problem I have been living with for years.

Have feedback?

If you spot an error, have an idea for a question we should add, or just want to tell me how it went, I want to hear about it. Email me