Kscope · Console · 2026-10-15
Opinion on the Console roadmap copy read on 2026-10-15. The page is the Oct 14 roadmap. This copy moves the assistant from last place to Wave 4 and slides the deal desk, sector depth, the people graph, and transcripts back one wave. Wave 3 is unchanged.
| Wave | This copy | The page before it |
|---|---|---|
| 3 | Gives, and one company on one record | Same |
| 4 | Ask the filings: assistant, full-text search, passage click-through, document compare, tear sheets, Excel add-in | Deal desk |
| 5 | Deal desk | Sector depth |
| 6 | Sector depth | People graph |
| 7 | People graph | Transcripts |
| 8 | Transcripts, waiting on the call text | Assistant, last because a question costs money |
The sentence that held the assistant until the data waves were done is gone. The scorecard matches the new numbers. The banner still names Wave 3 as next, which is right.
Gives, plus one company shown as one record.
That is the missing verb. After Wave 2 the graph covers who owns, funds, advises, lends, and votes. Wave 3 adds who gives, and stops the same company from appearing as three companies.
Show them together and leave the legal entities intact. The operating company, the holdco, and the former public parent each file. One page with three linked records keeps that trail. Pluralsight is the test.
Gives needs a minimum person merge before it ships. A grant tied to a split person lands on the wrong profile. The full "who knows whom" map can wait for Wave 7. The merge that Gives depends on belongs in Wave 3.
On the graph, yes. On the terminal, the three gaps the roadmap names stay open, and this copy changes only when the first of them arrives.
What is live covers the US filing wedge: institutional holdings, beneficial-ownership filings, insider transactions with plan sales split out, registered-fund holdings, fund votes, the short-interest print and the daily short-volume print, Form D, adviser filings, business-development-company loan pieces, material contracts, and the filing behind the number. Wave 2 put the new surfaces on that data: insider track record, funds managed, the credit monitor, the adviser screener, proxy voting, and fund votes.
Identity is the part Wave 3 starts and does not finish. The entity note of 2026-09-18 counts 234,632 Exhibit 21 subsidiary disclosures, 24,900 resolved subsidiary links, and 22,044 entities with an ultimate parent. Goldman is the one confirmed family, and most of its Exhibit 21 names are still unresolved. People are still split across records. The filed subsidiary list belongs on the company page once the lists stop truncating. Name resolution for every subsidiary can lag the list.
The roadmap's line that 672,000 private companies are mostly unprofiled counts shells and vehicles with operating companies. The operating cut, recounted on 2026-10-09, is 123,381 companies, 82,476 of them private. Wave 3 is a rollup of records already in the file.
Estimates, real-time and global prices, broker research, and ratings stay a license. That call on the roadmap is right. Training a model on the filings does not recreate someone else's unpublished estimates.
The assistant is the piece that belongs earlier. Both terminals shipped chat-with-a-citation in the last year, and Console already exposes the same numbers through tools. A buyer can be shown "ask the graph, open the filing" without waiting on league tables. Transcripts belong after it, as this copy has them. The assistant answers from filings and says when a call is not loaded.
Three limits keep the cost honest. The old page was right that a question costs money. Moving the wave up leaves that cost in place.
This copy disagrees with itself. The Wave 3 writeup still extracts deal values from the merger filing and keeps one row per deal. The scorecard now says deal values are Wave 5, because that cell moved with the deal desk. Wave 5 says it needs those values and the one-company record.
Leave the extraction in Wave 3. Wave 5 is the precedent screen, adviser fees, league tables, private-placement screens, sponsor portfolios from Form D, the adviser file, and the loan books, plus point-in-time comps and a valuation builder.
Two additions for that desk, both from filings already in hand:
If the team spends the next build on the assistant and the deal values slip out of Wave 3, Wave 5 has nothing to screen.
Putting the assistant at Wave 4 closes the "buyers expect a chat box" gap sooner, on US filings. Private-markets and deal-team buyers, two of the four audiences named on the roadmap, wait an extra wave for multiples. That is a fair trade when the next sale is a terminal demo. It is the wrong trade when the next sale is a deal team.
Check the roadmap's "~1,180 companies, including airlines and cruise" before those rows are shown. The industry-metric note through 2026-09-22 does not list airlines, cruise, telecom, retail, or hotels as live packs. Home builders are marked built there, and that note does not give a row count.
Insider reports outside the United States are public filings, one regime at a time, the same shape as the US form. The US file is classified and joined to a person: about 11.28 million Form 4 and Form 5 transactions, 19,910 issuers, 209,003 people, a clean panel from 2004. The roadmap files the other regimes under out of scope. Leave them there until a client asks, and then do one country. Canadian company filings continue through QuoteMedia. A Canadian insider report is its own regime.
Keep Wave 3 as Gives plus one page of linked records, and keep deal values inside it. Ship a capped assistant in Wave 4 that answers through the tools and opens the filing. Split search-to-the-sentence, the slide export, and the Excel sideload out of that wave. Leave the precedent screen, fees, and league tables in Wave 5, and turn clean sector packs on when the rows are fixed, ahead of Wave 6.