Consultant database management where you ask the agents, see the source behind every figure and approve before anything reaches a consultant
Every consultant profile current each quarter, every figure tied to its source.
See one case, screen by screen ↓Consultant Database Management is software that keeps every strategy profile current on every investment consultant database each quarter, from approved fund data: the consultant database updates an asset manager's RFP and consultant relations team carries. The coordinator works with seven agents in plain words — asks why a figure differs, sees the source behind every value, and approves or rejects each proposal with a reason. The agents plan the cycle, fill every field with its source, find conflicts across portals including wrong numbers already live, test gross and net performance presentation, draft notices and Japanese and German text, and submit each portal with a receipt and a read-back. Compliance approves every performance figure and Consultant Relations signs every message to a consultant.
AI in asset management: a buyer's guide to agentic AI for investment managers →
Updated
Consultant database management is the quarter-end work of keeping every strategy profile current in every investment consultant database: assets and accounts, returns gross and net of fees over standard periods, benchmark returns, risk statistics, holdings and characteristics, team, fees and commentary. Each portal has its own field names, precision, definitions and deadline — strategy assets or composite assets, effective or modified duration, composite or vehicle inception — and the RFP and database coordinators or consultant database analysts copy the numbers by hand from the performance system, factsheets and GIPS composite reports, then send everything to marketing compliance as a spreadsheet to re-check.
The risk is the number nobody compares. A value typed into the CRM or into one portal outside the cycle is never checked against the source, so a transposed figure can sit in front of a consultant for a whole quarter. Team bios go years out of date, a portfolio manager's departure reaches a consultant late, performance goes out gross without net, and when an examiner asks which source fed which performance figure, there is no record to show.
Estimated split for a two-person team keeping 60 strategies on 12 consultant portals, by hand and with the solution. By hand is our estimate (9,012 fields at about 40 seconds each is about 100 hours of keying); with the solution is the design target.
Seven agents source, reconcile, check and submit every profile; the coordinator releases, Compliance approves figures, Consultant Relations signs.
Quillfield Capital keeps 60 institutional strategies on 12 consultant portals: 300 profiles and 9,012 fields in the Q3 2026 consultant database updates. It is Thursday, October 15, business day 11. Here is how the coordinator works with the agents — asks, checks the evidence, approves — screen by screen, in the working solution.
Jamie's Today screen opens on four numbers — 40 of 300 profiles submitted, 8,798 of 9,012 fields sourced with 0 sent without a source, 13 of 48 conflicts open, 1,744 fields waiting for approval — and on what each agent did since yesterday, every line cited. Jamie types the question on his mind: why does Fenwick Hale show $4.21B? The Conflict agent answers in plain words with three sources — the CRM history, factsheet page 1 and the performance system — and a button that opens the conflict.
Conflict agent: “The performance system and the Sep 30 factsheet both say $4.12B. I changed nothing; the fix waits for Jamie Cole.”
Global Small Cap Equity has 320 fields across the 12 portals: 160 fresh, 146 ready for approval, 12 stale, 1 waiting and 1 conflict. Jamie clicks the red mark. Fenwick Hale shows $4,210.0M live, typed by hand on October 7 with no source; the approved source, the performance system, says $4,118.6M. The panel beside the grid also shows Fenwick Hale's own definition of the field and who approves it.
Portal asks: “Total strategy assets, USD millions, all vehicles and accounts.”
Any value opens its source. Here the performance system record for composite QF-C-114, period Q3 2026 locked October 9 at 6:40 p.m. by Tomás Ferreira: strategy assets $4,118.6M, composite assets $3,871.2M, 14 accounts in the composite, returns gross and net for each period, and the net-of-fee method. The Sourcing agent filled 319 of the strategy's 320 fields from records like this one, never from the CRM.
“Opened from the record · every value keeps its source, as-of date and version”
The conflict CF-0412 lays out every place the figure lives. The performance system and the 11 other portal drafts say $4,118.6M, and the September 30 factsheet says $4.12 billion: they match. Fenwick Hale's live value and the CRM product record differ. The composite report's $3,871.2M is not compared, because composite assets are a different field. The agent's fix has four parts: correct Fenwick Hale on release with a short correction note, keep the 11 drafts, ask the CRM record's owner to fix it, and send the figure to Compliance.
Jamie replays how the agent got there. It read strategy assets on all 12 portals, read the approved sources, matched Fenwick Hale's definition 2.1 to strategy assets, and traced $4.21B to a CRM record edited by hand on October 6 at 4:12 p.m. London with no linked source. It explained the $91.4M (2.2%) gap, found no RFP, DDQ or other portal that used the CRM figure, and escalated at 98% confidence.
“Cause: a hand-typed CRM edit swapped two digits (4.12 → 4.21), then the CRM figure was typed into Fenwick Hale's form outside the cycle. The agent changed no value.”
Jamie tests the reasoning in the conversation beside the conflict: could $4.21B include model assets? The agent answers no, and cites the firm's AUM policy and Fenwick Hale's question. Satisfied, Jamie approves the fix with reason code AP-02, source error corrected. The decision is signed with his name and role and recorded on the audit trail.
“Model-delivery assets for this strategy were $0.31B at Sep 30, and firm policy §3.2 reports them as assets under advisement, never as strategy assets. … And $4.12B + $0.31B is $4.43B, not $4.21B.”
Elena's queue holds the performance and AUM fields in strategy batches, each already tested by the Compliance check agent: net with gross for the same period, 1-, 5- and 10-year periods, benchmark named, composite disclosure attached. Clean batches can be approved together; flagged ones, like US High Yield with a benchmark revised October 8, open one by one. She approves the 145 performance and AUM fields for Global Small Cap. Jamie could not have approved them, and Elena cannot approve anything she prepared.
“Clean batches can be approved together; flagged ones open one by one. Nobody approves what they prepared.”
Jamie releases the corrected strategy to all 12 portals. The Submission agent files each in its own format — web form, spreadsheet template or data feed — and the marks on the grid turn to submitted for ten portals. Rheintal and Aoba stay queued for their German and Japanese reviewers.
“Submitted to 10 portals with receipts · 2 queued for reviewers.”
Every submission keeps its receipt and the result of reading each field back from the portal. Fenwick Hale's corrected profile is FH-Q3-26-0471, all fields match. When Wexmoor stored one decimal, Compliance accepted the two rounded fields with reason code OV-03. Aoba's sign-in expired overnight, so the agent paused: nothing was sent or lost, and Dana Whitaker renews the credential in the vault.
The audit trail shows the morning in order: the Conflict agent's escalation with its four sources, Jamie's approval with AP-02, Elena's approval of 145 fields against the marketing review policy, the release and the submission. It also shows the October 6 CRM edit and the October 7 hand entry — no reason, no approval. Every step in the cycle is hash-chained and kept 7 years, and Owen exports the evidence pack.
The cycle report: 48 conflicts caught by the agents, 0 figures sent without a source.
The same quarter, seen by the four people who carry it — what it looked like by hand, and what it looks like with the agents.
Seven agents source, reconcile, check and submit every profile; the coordinator releases, Compliance approves figures, Consultant Relations signs.
Builds the quarter's plan from each portal's deadline and each source's arrival date, assigns owners, watches the SLAs — every portal ready 3 business days before its deadline — and drafts escalations and extension requests.
Fills every field from approved sources only — performance system, portfolio accounting, firm AUM report, factsheets, composite reports, prospectuses, content library, fund administrator and market data — and stamps each with its source, as-of date and version.
Compares each field across every portal's live value, this quarter's drafts and the sources; matches each portal's definition so different fields are left alone; traces the cause and proposes a fix with its confidence.
Tests every performance figure against the presentation rules in the firm's marketing review policy — net with gross for the same period, 1-, 5- and 10-year periods, benchmark named, composite disclosure attached, as-of date — and builds the reviewer's packet.
Watches the as-of date of every source and the HR, product and ESG records; raises stale and change alerts, such as team bios over 12 months or a co-PM's resignation, and drafts the notice to every consultant that carries the strategy.
Drafts commentary, team and change text from approved English only, in each portal's language — Japanese and German — with the original beside it and names checked against the firm's register.
Formats each portal's update as a web form, spreadsheet template or data feed, submits released batches, captures the receipt, reads every field back from the portal and drafts correction notes.
Maker, checker, signature. The agents propose; a named person decides.
Every screen has an ask box: Today, each conflict, each review queue, submissions and the audit trail. The agents answer in plain words, cite the record behind the answer and open the screen where you act.
Why does Fenwick Hale show $4.21B?
Because a CRM record was edited by hand on Oct 6 (4.12 → 4.21) and typed into Fenwick Hale's form on Oct 7, outside the cycle. The performance system and the Sep 30 factsheet both say $4.12B. I changed nothing; the fix waits for Jamie Cole.
Which portals are at risk this week?
One: Fenwick Hale, due Thu Oct 22, with Global Small Cap held behind CF-0412 and the EM Local Debt notice unsigned. The SLA wants it ready by Mon Oct 19. Corvane (Oct 23) is on track once Compliance clears 52 profiles. Pellham needs an extension for two private credit funds whose marks arrive Nov 12.
What still needs Compliance?
1,296 performance and AUM fields across 60 strategy batches, one per strategy. Three carry a flag: Global Equity (Calloway's template has no net column), US High Yield (benchmark revised Oct 8) and Direct Lending IV (net IRR waits for Q3 marks). Global Small Cap is held behind CF-0412; the other 56 pass every presentation check.
How sure are you?
Two approved sources agree to the tenth of a million, the definitions match, the CRM change has no linked source, and the swapped digits explain the exact gap of $91.4M (2.2%). That is why I am at 98%. Below 90% I would have escalated without a proposal.
13 screens from the working solution, on its sample data. Pick one to see it large.
Thresholds and definitions are versioned — version 14 is live since October 1 — and a change waits for a second approver in Compliance.
| Setting | Default | Choose from |
|---|---|---|
| Sourcing agent fills a field at or above | 95% confidence | A percentage the firm sets · below it, the field waits and asks its owner |
| Conflict agent proposes a fix at or above | 90% confidence | A percentage the firm sets · below it, the agent escalates without a proposal |
| Compliance check flags below | 97% confidence | A percentage the firm sets |
| Team bios are stale after | 12 months | In months · firm diversity data has its own limit, also 12 |
| ESG policy is stale after | 12 months | 18 months proposed as version 15, pending Compliance's approval |
| Values match when | Equal after rounding to the portal's precision | Set by the firm |
| Currency conversion | London 4 p.m. fixing, last business day of the quarter | Set by the firm |
| CRM | Reference only: never a source for numbers | Locked |
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 (about five weeks by hand: 9,012 fields at about 40 seconds each, about 100 hours for two people who also answer RFPs) · “typical” = published figures (Cerulli Associates via Institutional Investor — consultants have intermediated more than 60% of institutional flows; in 2019, 91% of managers found it challenging to meet RFP and DDQ deadlines). Rules named: SEC Rule 206(4)-1 (Marketing Rule; compliance date Nov 4, 2022) and staff FAQ; SEC Rule 204-2; GIPS. People, firms, consultants, portals and strategies named on this page are fictional — characters and sample data in the working solution.
Consultant Database Management is software that keeps every strategy profile current on every investment consultant database each quarter, from approved fund data: the consultant database updates an asset manager's RFP and consultant relations team carries. The coordinator works with seven agents in plain words — asks why a figure differs, sees the source behind every value, and approves or rejects each proposal with a reason. The agents plan the cycle, fill every field with its source, find conflicts across portals including wrong numbers already live, test gross and net performance presentation, draft notices and Japanese and German text, and submit each portal with a receipt and a read-back. Compliance approves every performance figure and Consultant Relations signs every message to a consultant.
Consultant database maintenance starts from a calendar. The Planning agent builds the quarter from each portal's deadline and each source's arrival date, and the Sourcing agent fills every field as data is final — the performance system once the period is locked, holdings from portfolio accounting, factsheets and composite reports from document management. A field whose source is not final waits and asks its owner, and the Freshness agent flags stale bios, policies and diversity data before a consultant sees them.
Where a database profile is an advertisement under SEC Rule 206(4)-1, gross performance must be shown with net performance over the same period and with the same methodology, and the performance of a portfolio other than a private fund needs 1-, 5- and 10-year time periods. Your compliance team decides when a profile is an advertisement; the Compliance check agent tests every figure against those points and your marketing review policy — net with gross, the periods, the benchmark named, the composite disclosure, the as-of date — and hands the reviewer a packet. It never approves.
The Conflict agent reads each portal's live value alongside this quarter's drafts and the approved sources, matches the portal's definition so composite assets are not compared with strategy assets, and traces the cause — in the demo, a hand-typed CRM edit with two digits swapped. It proposes a fix with its confidence and never overwrites a value; the coordinator approves it with a reason code.
Yes. Every field keeps its source, as-of date and version, and every approval, release and submission is on a hash-chained audit trail with the evidence and reason code. Internal Audit or Compliance exports an evidence pack per cycle, strategy or portal, and records are kept 7 years, longer than the 5 that Rule 204-2 asks for.
No. The agents prepare, check and explain; they never approve, overwrite a value, take a number from the CRM or send a message to a consultant. A batch releases only when every performance and AUM figure is approved by Compliance, nobody approves what they prepared, and the Submission agent submits only released fields, using credentials from the vault.
Yes. The Narrative agent drafts commentary and change text from the approved English in each portal's language, with the original beside it and names checked against the firm's register. A native-language reviewer approves every non-English text, and those portals stay queued until they do.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your marketing review policy, field definitions, last quarter's portal answers and source systems — and it goes live once every quality gate has passed. We will walk you through it on your own strategies first.
Finals decks the same day, with every number stamped to its source.
1-min demoInstitutional Client OnboardingWon mandates funded in weeks, not months, with every signer and bank detail proven.
1-min demoClient Inquiry ManagementClient requests answered from the source systems, every figure cited and cleared.
1-min demoWe'll run Consultant Database Management on a sample of your own strategies and last quarter's portal answers.