Fiduciary Engine AI-assisted Form 1041 prep for estate & trust firms

Turn a pile of K-1s, 1099s, and a trust instrument into a filing-ready 1041 + beneficiary K-1s + state fiduciary returns — with a CPA/EA signing off.

● CONCEPT / PROTOTYPE — internal

The wedge: Form 1041 work is high-volume, low-glory, deadline-driven. Estate firms either farm it to a slow $400–600/return CPA, or grind it in-house and lose money. It's repetitive, rules-based, and document-driven — perfect for automation. We're not competing with TurboTax. We're competing with "a paralegal + a $400 CPA."

How it works — click Run to walk a sample estate through the engine

Intake

K-1s, 1099-INT/DIV/B, brokerage composites, prior 1041, trust instrument

✓

Classify

Estate vs. trust · simple vs. complex · grantor status · fiscal/calendar year

✓

Compute (core IP)

DNI · income-distribution deduction · tier-system allocation · Schedule B

✓

Generate

Filing-ready 1041 + per-beneficiary K-1s + state fiduciary returns

✓

Review & sign

CPA/EA reviews flagged fields and signs off — we never auto-file

✓

Sample: The Harding Family Trust · complex trust · calendar year · 3 beneficiaries

Distributable Net Income (DNI)

$84,200

interest + dividends + net rental, less tax-exempt

Income Distribution Deduction

$72,000

lesser of DNI or distributions

Trust taxable income

$12,200

retained at trust level

Schedule K-1 allocation — character flows to beneficiaries by the tier system

Beneficiary	Ordinary income	Qualified div	Net LT cap gain	Total K-1
A. Harding (1st tier)	$18,400	$9,600	$0	$28,000
B. Harding (1st tier)	$18,400	$9,600	$0	$28,000
C. Harding (2nd tier)	$10,520	$5,480	$0	$16,000
Distributed to beneficiaries	$47,320	$24,680	$0	$72,000

Reviewer flags — surfaced for CPA/EA before sign-off

low confidence Brokerage 1099-B wash-sale adjustment extracted at 71% — verify against the composite statement.
classification Capital gains allocated to trust per instrument §4.2 default — confirm no discretionary distribution of corpus this year.
passed DNI, income-distribution deduction, and all three K-1 character lines reconcile to the as-filed prior year within tolerance.

All figures above are illustrative placeholder data for a fictional trust — not a real return.

Why this works

The moat

DNI / allocation engine — the actual tax brain (tier system, cap-gains allocation, simple vs. complex). Get it rock-solid → defensible.
State fiduciary returns — 40+ states, each different. Incumbents are weak here. Start with our states.
E-file (MeF) — needs an EFIN / authorized provider. Real gate = real barrier to copycats.

Business model

Per-return, service-led: $150–300/return — undercut the CPA, fat margin once automated.
Seat + volume tiers for firms doing 50–500 returns/yr.
B2B only — estate/trust firms, family offices, small CPA shops. No consumers.

Why us

Tom runs a multi-state firm + Estate Done Right — design partner AND first customer.
Same proven pattern as BillForensics / Trust Health: docs in → rules engine → reviewable output.
We dogfood on our own estate matters before selling a single seat.

Path to revenue

DNI/allocation engine + federal 1041 + K-1 generation, validated against 10–20 of our own as-filed returns (golden set).

PA + NJ state fiduciary returns.

Service offering live — we prep, AI does ~90%, a CPA signs. Sell outcomes, not software. Funds the build.

2–3 external pilot firms.

Self-serve SaaS + EFIN e-file.

Open decisions

Product name — FiduciaryAI / TrustReturns / 1041.ai
Entity — net-new vs. roll under Estate Done Right
CPA/EA partner — who's the sign-off authority?
v1 states — default PA + NJ
Golden test set — pull 10–20 completed 1041s + K-1s from the firm book (Day-1 blocker)

Concept page · stood up temporarily on petre11i.ai · source docs in fiduciary-engine/