K-1 aggregation that tracks every expected K-1, cites every box and catches every estimate swing before export
Every K-1 tracked, read and cited, and every estimate swing caught before export.
See one case, screen by screen ↓K-1 aggregation is collecting every Schedule K-1 and K-3 an investor receives from the partnerships it holds, tracking which are still missing, and extracting them into tax-ready data. K-1 Aggregation brings Schedule K-1 automation to that work with seven agents: they gather each K-1 from GP portals, email and file drops, read every box and footnote with its page cited, explain estimate swings, tie out to the books and follow up on late K-1s. An analyst reviews and the tax manager approves the export to tax software.
AI in asset management: a buyer's guide to agentic AI for investment managers →
Updated
K-1 aggregation is the work of getting every Schedule K-1 an investor is owed, from every partnership it holds, into tax-ready data. For an LP with hundreds of private fund holdings that means a K-1 from each GP, often with a Schedule K-3, a March estimate, a state schedule and sometimes an amended K-1, arriving through hundreds of GP portals, a shared mailbox and fund administrators' file drops. The investor tax team logs in portal by portal, re-keys every PDF into spreadsheets and the tax software, and keeps a list of what is still missing by hand.
The risk sits in the footnotes and the calendar. UBTI, Section 199A, state sourcing and nonresident withholding live in supplemental statements, not boxes, and UBTI is often given as a percentage of another line. On extension, K-1s can arrive as late as September 15, the same day a fund of funds' own extended return is due, and the final can move far from the estimate in amount and in character: ordinary income turns into a loss, a new UBTI line appears. For a fund of funds, every late or changed underlying K-1 means late or amended K-1s to its own partners.
Estimated split for one K-1 in season at a large LP, by hand and with the solution.
Seven agents gather, read, compare and follow up on every K-1; an analyst reviews and the tax manager approves.
Quillfield Private Markets Partners IV, a fund of funds filing Form 1065, must issue its own K-1s to 1,912 partners by Tuesday, September 15. It is Tuesday, September 1. Here is how the investor tax team works with the agents, screen by screen, in the working solution.
Hana's home screen shows the season: 2,512 K-1s expected, 2,031 finals in, 1,812 reviewed, 1,640 exported to tax software. Under “Needs you” sit a swing at Harlow Growth Partners III, a UBTI footnote at Tamsin Bay that needs her judgment, an amended K-1, a tie-out break and a French statement. Beside it, the agents report what they did overnight and she can ask them anything in plain words.
Estimate agent: “One swing needs a person: Harlow Growth III’s Box 1 moved from $95,000 to −$182,400. Footnote 4 explains it.”
QPMP IV's roster is built from its holdings in investment accounting: 214 K-1s expected, 171 finals in (80%), 25 estimate only, 18 nothing yet, and 14 days to its Form 1065. Each row tracks estimate, final, extracted, reviewed and exported, with the agents' note and the GP's promised date. Sablewood Energy Partners IV, Pinecrest Hollow Fund of Funds III and Marlowe Infrastructure Partners II sit at the top.
Sablewood: “GP extended; promised Sep 14, one day before the deadline. Propose estimate as placeholder.”
The Collector agent pulled Harlow Growth Partners III's final 2025 K-1 from the GP portal, 4 pages with the K-3 attached, and matched it by partner ID. The Extraction agent read it in 11 seconds: 15 boxes, Items J, K and L, 5 state lines and K-3 Parts II and III, at 99% confidence. The Footnote agent mapped 8 footnotes, including UBTI of −$41,800 in Box 20 code V, the Section 199A statement and California withholding of $2,236.
Every value sits beside the read-only original, with a number that opens the passage it came from.
Box 1 fell from $95,000 on the March 12 estimate to −$182,400: footnote 4 is QPMP IV's 1.8473% share of a $9,874,000 ordinary loss at Tidewater Cold Chain Holdings from its December 18, 2025 recapitalization. Box 9a rose $202,300 on footnote 5, the November 24 sale of Orrin Medical Devices. The agent shows what it changes in QPMP IV's own Form 1065 draft, which carried the estimate as a labeled placeholder in export EXP-0827-03.
“Net income moves only −$90,240, but its character flips from ordinary income to ordinary loss, and UBTI of −$41,800 is new.”
Hana asks what happens if they keep the estimate. The answer: QPMP IV's return would overstate ordinary income by $277,400 and understate long-term gain by $202,300, its 214 tax-exempt partners would miss the UBTI, and amended K-1s for 1,912 partners would follow. She accepts with AC-02 · “Final supersedes estimate; swing explained by GP footnote”, and a revised export goes to the tax manager.
“A revised export for QPMP IV goes to the tax manager; you cannot approve it yourself.”
Revised export EXP-0901-02 shows what changes in the tax software: Box 1 from $95,000 to −$182,400, Box 9a from $410,000 to $612,300, and new lines for UBTI (−$41,800) and foreign taxes ($3,140). The Export agent ran its 12 checks, from “Every line comes from a reviewed K-1” to “Preparer is not the approver”. Elena approves with AP-01 · “Reviewed with sources”, 14 days before the deadline.
“Analysts prepare, the tax manager approves; nothing reaches the tax software without two people.”
Tamsin Bay Real Estate Partners V gives the debt-financed share as 37.2% of net rental income “before depreciation”. Under section 514 the share applies to income and its deductions, which points to Box 2 after depreciation; the GP's wording points the other way. Below its 90% line, the agent puts both readings to Hana: Box 2 of $286,400 gives UBTI of $106,541, before depreciation $412,900 gives $153,599, or she can ask the GP.
The Footnote agent never guesses a percentage; below 90% confidence it waits for a person.
Fonds Infrastructures Laurentides, S.E.C. reports on a Relevé 15 in French. The agent reads the original, shows it beside an English workpaper and converts at the Bank of Canada 2025 average, CAD 1 = USD 0.7300: net business income of CAD 412,880 becomes $301,402, and CAD 41,290 withheld on distributions maps to a foreign tax credit of $30,142. Every figure is cited to the French page, and the workpaper notes that QPMP IV reports this foreign partnership on Form 8865.
Brookhollow Buyout Fund VIII reports $1,245,000 of distributions in Box 19 A; investment accounting booked $1,195,000. Footnote 8 says $50,000 was declared December 30, 2025 and paid January 5, 2026, so the custodian booked it in 2026. Contributions, beginning capital against the 2024 K-1, partner ID and the 2.0402% share all tie. Hana can accept with a reconciling item or ask the GP to confirm.
The Tie-out agent reads investment accounting and never posts to the books.
The Follow-up agent counts back from each investor's own deadline: reminders at 45, 30 and 21 days, an escalation to the GP's head of investor relations at 14, a placeholder plan at 7. For QPMP IV that is today. Pinecrest's promise of Sep 25 is after the deadline, so it proposes the prior-year ratio as a labeled placeholder and an amended return plan; Marlowe's three unanswered reminders become an escalation copied to Tomás. He approves the drafts before anything is sent.
The Follow-up agent writes only to GP contacts on file; every escalation needs the operations manager.
The same tax season, seen by four people at Quillfield Capital: what it looked like by hand, and what it looks like with the agents.
Seven agents gather, read, compare and follow up on every K-1; an analyst reviews and the tax manager approves.
Gathers K-1s, K-3s, estimates and amended K-1s from 312 GP portals, the K-1 mailbox and 14 administrators' file drops, matches each to its holding and removes duplicates.
Reads every box, code, Items J, K and L, every K-3 part and state schedule, in any language, and cites the page for each value.
Turns supplemental statements into tax codes: UBTI and debt-financed ratios, Section 199A, state sourcing and withholding.
Compares each final with its estimate box by box, flags swings over the threshold, traces each to its footnote and shows what it changes in the investor's own return.
Ties Box 19 distributions, contributions and capital to investment accounting and to last year's K-1, and turns a break into a GP query with its evidence.
Plans follow-ups back from each investor's deadline, drafts reminders and escalations, and reads GP replies into promised dates.
Builds the tax software import with activity numbers, runs its 12 checks, labels placeholders for amendment and assembles the evidence pack.
Review, approve, message GPs. The agents propose; a named person decides.
Today, the roster and every K-1 have a place to ask about a holding or tell the agents what to do. Answers point to their source; people decide.
Which K-1s could miss September 15?
QPMP IV has 18 K-1s with nothing yet, 14 days out. At risk: Sablewood Energy Partners IV, promised Sep 14; Marlowe Infrastructure Partners II, no reply to three reminders; Pinecrest Hollow Fund of Funds III, promised Sep 25. I have drafted escalations and placeholder plans for approval.
Why is Box 1 a loss?
Box 1 is −$182,400 because footnote 4 allocates QPMP IV 1.8473% of a $9,874,000 ordinary loss at Tidewater Cold Chain Holdings, LLC: transaction bonuses and debt extinguishment costs on its Dec 18, 2025 recapitalization. 1.8473% × $9,874,000 = $182,402.
What happens if we keep the estimate?
If QPMP IV files with the estimate, its return overstates ordinary income by $277,400 and understates long-term gain by $202,300, and its 214 tax-exempt partners miss −$41,800 of UBTI. An amended return and amended K-1s for 1,912 partners would follow. I recommend accepting the final.
Which partners get the UBTI?
The −$41,800 of UBTI passes through QPMP IV to its 214 tax-exempt partners by capital share. It appears on their QPMP IV K-1s in Box 20 V; taxable partners are not affected.
13 screens from the working solution, on its sample data. Pick one to see it large.
The rules every K-1 is checked against are settings: the administrator proposes, the tax manager approves, and every version is logged.
| Setting | Default | Choose from |
|---|---|---|
| Swing threshold | $50,000 and 25% on a box | Any dollar amount and share of the estimate |
| Review below confidence | 95% | Boxes below this are highlighted for the analyst |
| Stop and ask below | 90% | Footnote mappings below this wait for a person's decision |
| Always flag | New UBTI, new K-3 items, character changes | Flagged whatever the amount |
| Bulk review allowed when | 97% or more, tie-out passed, no flags | Clean K-1s only |
| Placeholder for a missing K-1 | GP estimate, else prior-year ratio, always labeled | Listed in the export for amendment |
| Distribution tie-out tolerance | $1,000 | Any dollar amount |
| Follow-up cadence | Reminders at 45, 30 and 21 days; escalation at 14; placeholder plan at 7 | Counted back from each investor's extended deadline |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (IRS Partner's Instructions for Schedule K-1 (Form 1065), 2025, on the September 15 extended deadline; eqvista, “Why Schedule K-1 is late”, on 30–45 days to close the books and LPs waiting nearly nine months). Quillfield Capital and every fund, GP, person and amount named on this page are fictional — characters and sample data in the working solution.
K-1 aggregation is collecting every Schedule K-1 and K-3 an investor receives from the partnerships it holds, tracking which are still missing, and extracting them into tax-ready data. K-1 Aggregation brings Schedule K-1 automation to that work with seven agents: they gather each K-1 from GP portals, email and file drops, read every box and footnote with its page cited, explain estimate swings, tie out to the books and follow up on late K-1s. An analyst reviews and the tax manager approves the export to tax software.
The Collector agent signs in to the GP portals with credentials kept in a vault, checks the K-1 mailbox and the administrators' file drops, and matches each document to its holding on the roster. Duplicates are kept once and unmatched files go to a person. When a portal refuses sign-in, it asks operations to renew the credentials; it never accepts new portal terms or changes MFA itself.
The Estimate agent compares the final with the estimate box by box. A box that moves more than the threshold, $50,000 and 25% by default, or any new UBTI, new K-3 item or change in character, is flagged and traced to the footnote behind it, with what it changes in the investor's own return. The analyst accepts the final with a reason code and the tax manager approves the revised export.
The Footnote agent reads the supplemental statements and maps them to tax codes: UBTI and debt-financed ratios, Section 199A, state sourcing and withholding. Each mapping is cited to its footnote with a confidence. Below 90% it stops and asks the analyst, as with a UBTI base given as 37.2% of net rental income “before depreciation”.
Yes. K-3 data extraction runs with the K-1: the Extraction agent reads the K-3 parts, such as Parts II and III on the sample K-1, and cites each item. A missing K-3 can be requested from the roster, and the export flags K-3 items the investor needs for its own K-2 and K-3.
The Follow-up agent plans reminders back from each investor's own extended deadline, reads GP replies into promised dates and drafts escalations for the operations manager to approve. When a promise lands too late, it proposes the GP's estimate or the prior-year ratio as a labeled placeholder and an amended return plan. An amended K-1 supersedes the original, shows the difference to what was exported and flags a re-export.
Yes. The agents never accept a final, send an export, change the roster or message a GP outside the approved template. The analyst reviews and prepares, the tax manager approves every export, the operations manager approves every escalation, and every decision is signed with a reason code on a hash-chained audit trail.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your investor entities, holdings, deadlines and a sample of past K-1s — and it goes live once every quality gate has passed. We will walk you through it on your own K-1s first.
Every KPI cited to its page, every covenant tested the day the pack lands.
1-min demoPrivate Equity Deal ScreeningEvery CIM tested against the fund’s own limits in minutes, with every page cited.
1-min demoVirtual Data Room Due DiligenceEvery document read and every red flag cited to its page, before you sign.
1-min demoWe'll run K-1 Aggregation on a sample of your own K-1s, estimates and statements.