botminds.aiRequest a demo
SystemCompliance & Ethics

Speaker Program Sentinel

Speaker program compliance monitoring — every program scored on the OIG’s eight red flags, before and after it runs

Every speaker program scored against the OIG's eight red flags, before and after it runs.

Watch the filmBook a demo
See one case, screen by screen ↓
target100%of programs scored on all eight OIG red flags
demo7of 8red flags found on one steakhouse dinner in Tulsa, each with the line it came from
demo86 → 52risk score once the program moves to an office with no alcohol
target12of 18unannounced audits a month aimed by the risk score
The problem

Why most speaker programs are never really checked

The OIG’s 2020 Special Fraud Alert names the signs of a speaker program that is a kickback: a restaurant or entertainment venue, no new medical information, alcohol or a meal above modest value, many programs on one topic, the same attendees again and again, guests with no business reason, speakers picked by sales, and pay above fair market value. Every one of those can be checked. The trouble is where the evidence sits — the request in the events platform, the RSVP list, a handwritten sign-in sheet, the venue invoice, the nomination form, the speaker agreement, the label history.

So monitors sample. A few programs get a live visit, the rest are read after the fact, if at all. The repeat attendee only shows up when someone lines up twelve months of sign-in sheets by NPI. The two speakers who attend each other’s dinners, the wine on the invoice, the “guest” who shares the speaker’s surname — they are in the documents, but nobody has the hours to find them before the next program runs.

estimated≈3%of programs seen live by a monitor today
typical8red flagsnamed in the OIG alert — each to be checked on every program
Where the hours go on one high-risk programestimated
By hand10 days
With the solution1.5 days
  • Reading the request, venue and meal plan1 → 0.1 d
  • Checking attendee history across sign-in sheets2.5 → 0.3 d
  • Checking the deck and the label history0.5 → 0.1 d
  • Reviewing the nomination, fee and annual cap1 → 0.2 d
  • Reconciling sign-ins, RSVPs and the invoice2 → 0.3 d
  • Writing the monitoring report3 → 0.5 d

Estimated hours for one high-risk program, checked by hand and with the solution.

How it works

How a program moves

Seven specialist agents read, check, score and report on every program; a compliance monitor decides to clear, change, audit or refer.

What comes in
Programs inPlans, sign-ins and invoices · from the events platform
Agents at work
Program & Sign-in Readervenue · RSVPs · cost
Then
Venue & Meal Checker
Repeat Attendee Tracker
Topic & Label Checker
Speaker Selection & Pay Reviewer
Then
Risk Scorer & Audit Planner0–100 with reasons
Then
Monitoring Report Writerafter the event
A person decides
Compliance monitorclear · change · audit · refer
What comes out
Risk score, every flag cited
Unannounced audit plan
Signed monitoring report
One case, step by step

One dinner program, from request to signed report

A Corventa heart-failure dinner is booked at a Tulsa steakhouse for Friday, Oct 9 — two days away. Here is what the monitoring lead sees and decides, screen by screen, in the working solution, and how a program that already ran is closed out.

  1. 01Morning

    Every program in her region, already scored

    Lena Ortiz · Speaker program monitoring lead

    Lena opens the Sentinel board: 24 of 24 programs scored, before and after the event, across 7 territories. Three need her decision, two monitoring reports are ready to draft because the sign-ins and invoices are already read, and six unannounced audits are planned with zero advance notices. The board can be filtered by any of the eight red flags.

    Needs you: “SP-24117 · The Ember Room, Tulsa — Dr. Raymond Kessler · Fri Oct 9 · 7 flags.”

  2. 02Opened

    Scored 86 — high risk, 7 of 8 flags

    The agents

    She opens SP-24117. The agents read the request and run each check in view: the venue is a restaurant, the meal is over the rule, 6 repeat attendees among 14 invitees matched by NPI, 412 programs on deck CRV-HF-03 with the last label change 19 months ago, sales language in the nomination, cap use at 78 %. The score is the sum of the flags: venue +18, alcohol and meal +16, repeat attendees +16, sales-driven choice +12, same topic +10, no new information +8, pay near cap +6.

    Checked and clear: attendees with no business reason.

  3. 03Evidence

    Every flag, cited to the line it came from

    Speaker Selection & Pay Reviewer

    Each flag carries its citation and the rule behind it — the OIG factor, the PhRMA Code section or the company rule. Click the sales-driven flag and the nomination form opens beside it, with the sentence highlighted. Dr. Kessler has also been paid $39,000 — 78 % of the $50,000 annual cap — with 2 more programs booked (88 %). Lena can confirm or dismiss each flag.

    “Dr. Kessler is a top-decile Corventa writer in OK-114, with strong ROI on his programs — 4 new writers after his September dinner.” — nomination by Jordan Blake, sales representative

  4. 04Decision

    Change the venue, not just note the risk

    Lena Ortiz · Speaker program monitoring lead

    Her choices are clear, hold, refer, change venue or plan an unannounced audit. She asks the bureau to change the program: an office or hospital venue with no alcohol, or a virtual format. The message to Sam Patel at the speaker bureau is already written, with the reason — OIG factors 1 and 3 — and the expected score after the change: 52, from 86.

  5. 05Two days out

    Moved to a clinic — and still audited

    Lena Ortiz · Speaker program monitoring lead

    The program moves to Kessler Heart Clinic, $38 a head, no bar. The venue and meal flags clear; repeat attendees, sales-driven choice, same topic, no new information and pay near cap remain. So Lena also plans an unannounced live audit for Oct 9, with herself as the monitor. The “no advance notice to the field” switch is locked on.

    “Sales, marketing and the bureau never see the audit plan.”

  6. 06The pattern

    Two speakers taking turns

    Repeat Attendee Tracker

    Behind the repeat-attendee flag is a network built from twelve months of sign-in sheets and RSVP lists. Dr. Kessler and Dr. Brandt each attended 5 of the other’s Corventa programs this year — and Dr. Brandt is on the Oct 9 list again. Five regulars came to 4 or more of Dr. Kessler’s programs on one topic; all five are invited again.

    Line thickness = same-topic programs attended · red = 3 or more.

  7. 07After the event

    A handwritten sign-in sheet, read

    Program & Sign-in Reader

    SP-23988, a Veltrimab dinner at Harbor Grill in Baltimore, ran on Oct 2. The scanned sign-in sheet and the venue invoice are read row by row: 11 people signed in for 6 RSVPs, 3 without an NPI — one signed as “guest” and shares the speaker’s surname. The invoice comes to $138 a head against the $95 cap and includes two bottles of wine for $184. Score 62, 4 of 8 flags.

    No NPI invented — blank stays blank.

  8. 08One click

    The monitoring report, drafted with its sources

    Monitoring Report Writer

    Lena presses “Draft monitoring report”. The sign-in rows are reconciled against the RSVP list, the HCP master and the meal cap, and the report is written in four sections — program facts, attendance reconciliation, findings against policy, recommended actions — from 7 documents, every finding cited.

    “Every finding cites its source · intent is not concluded.”

  9. 09Signed

    She reviews, signs, and the program closes

    Lena Ortiz · Speaker program monitoring lead

    The recommended actions are specific: refer the “guest” who shares the speaker’s surname to investigations, recover the $184 wine charge and coach the representative on the no-alcohol rule, stop using Harbor Grill for Veltrimab programs and move the Oct 13 program there to an office or virtual format, and exclude two repeat attendees from same-topic invitations for six months. Lena can edit, sign, or sign and refer. She signs; the program closes.

Who it’s for

Built for everyone who touches a speaker program.

The same programs, seen by the monitor, the investigator, the bureau and the chief compliance officer — what their week looked like, and what it looks like now.

LO
Lena OrtizSpeaker program monitoring lead
Monitor
Before
Picks a handful of programs to visit and reads the rest after they have run.
Now
Starts each morning with every program scored, the flags cited, and the decision bar in front of her.
DO
Dana OkaforCompliance investigator
Investigator
Before
Gets a referral and starts by rebuilding the attendance history from sign-in sheets.
Now
Receives the evidence pack — the flags with citations, the documents and the attendance network.
SP
Sam PatelSpeaker bureau manager
Bureau manager
Before
Learns a program was a problem after it has run.
Now
Gets the change or hold request before the event, with the reason — and never sees the audit plan.
PR
Priya RamanChief compliance officer
Chief compliance officer
Before
Hears about speaker program risk one investigation at a time.
Now
Sees every US program scored, flags by OIG factor and by territory, and approves every change to the scoring rules.
Built on the engine

7 agents. Each with one job, and hard limits.

Seven specialist agents read, check, score and report on every program; a compliance monitor decides to clear, change, audit or refer.

Program & Sign-in Reader

Reads each program request from the events platform, and the scanned sign-in sheets and venue invoices after the event — venue, format, deck, RSVPs, cost per head, bar lines, and every sign-in row with name, NPI, signature and notes.

  • Cites the page and line for every value it extracts
  • Never invents an NPI — a blank stays blank
Venue & Meal Checker

Classifies the venue, compares the cost per head with the cap, and finds alcohol, a bar or a wine pairing in plans and invoices.

  • Quotes the invoice or request line for every flag
Repeat Attendee Tracker

Matches every invitee and signer to the HCP master by NPI, counts same-topic programs per person over 12 months, and spots speakers who attend each other’s programs.

  • A match below 0.9 goes to a person
  • Cites the sign-in sheet for each attendance
Topic & Label Checker

Counts programs per deck in the last 12 months and reports how long since the product’s last substantive label change, and what it was.

  • Uses only approved deck versions
Speaker Selection & Pay Reviewer

Reads nomination forms and territory plans for language that ties a speaker to prescribing, revenue or ROI, and checks the fee against fair market value and year-to-date pay against the annual cap.

  • Quotes the exact phrase that triggered the flag
  • Never reads prescribing data beyond the plan text
Risk Scorer & Audit Planner

Combines the eight factors into a 0–100 score, one line per factor with its citation, and proposes unannounced audits — every high-risk program first, then a random sample.

  • A person decides every high-risk program
  • Audit plans are visible to compliance only
Monitoring Report Writer

Drafts the monitoring report after the event — program facts, attendance reconciliation, findings against policy and recommended actions — in plain, factual sentences.

  • No finding without a citation
  • Recommends actions; never concludes intent
  • Draft only — a monitor signs
Compliance monitor

Decides; investigators take referrals. The agents propose; a named person decides.

Ask in plain words

Ask about any program, speaker or attendee

Monitors ask in plain words — or tell it what to change. Answers point to the request, sign-in sheet, invoice or nomination they rest on.

Why is SP-24117 high risk?

SP-24117 scores 86 — high risk. 7 of the OIG’s 8 red flags are present: a restaurant venue (The Ember Room steakhouse, private dining); $118 per head against the $95 cap, with a wine pairing requested; 6 of 14 invitees are repeat attendees, including speaker Dr. Brandt; the nomination says “top-decile writer … strong ROI”; and 412 programs ran on CRV-HF-03 with the label 19 months old. The event is in 2 days. Moving it to an office venue with no alcohol drops the score to about 52.

Which speakers take turns attending each other’s programs?

One pair: Dr. Raymond Kessler and Dr. Neil Brandt, both paid Corventa speakers. This year each attended 5 of the other’s programs on the same deck — Dr. Brandt signed in at Dr. Kessler’s Oct 1 dinner and Dr. Kessler at Dr. Brandt’s Sep 29 program. Dr. Brandt is invited again on Oct 9.

Plan unannounced audits for every uncovered high-risk program

Done. Two unannounced live audits added: SP-24117 · The Ember Room on Oct 9 (score 86) and SP-24142 · Riverside Country Club on Oct 15 (score 76). Every high-risk program in the next three weeks is now covered. Nobody in the field is told.

Lower the standard meal cap to $85 per head

Done. The standard-area meal cap is now $85 per head, was $95. All 24 programs re-scored: SP-24142 · Riverside Country Club is now also flagged for its meal at $92 per head, and its score rises from 76 to 86. The change is in Settings and in the audit trail, and the chief compliance officer is asked to approve it.

Every screen

The working solution, as it ships.

13 screens from the working solution, on its sample data. Pick one to see it large.

Sentinel boardEvery speaker program in the region, before and after it runs — scored, filterable by OIG red flag, with what needs a decision and the audits coming up.
Every programEach program with its venue type, risk score, red flags 1–8, stage and status — filterable by brand, risk, flag and status.
Program reviewThe risk dial, the points per OIG factor and how the agents scored it, with the decision bar: clear, hold, refer, change venue or audit.
Cited flagsEach flag with its citation and rule, and the source document open beside it — here the nomination with the sales language.
Change the programThe request to the speaker bureau, written with the reason and the expected score after the change.
Plan an unannounced auditLive or virtual, the monitor, the risk-based reason — and no advance notice to the field, locked on.
Repeat attendeesWho keeps coming back to the same topic, and which speakers attend each other’s programs — twelve months of sign-ins and RSVPs.
Sign-in sheet, readThe handwritten sheet read row by row, beside the flags it raised: guests without an NPI, wine on the invoice, a meal over the cap.
Attendance reconciliationRSVPs against sign-ins, people without an NPI, repeat attendees and food and beverage per head — each line cited.
Monitoring report, signedFindings against policy and recommended actions, signed by the monitor; the program closes, or goes to investigations.
Audit planThree weeks of unannounced live and virtual audits — risk-based, topped up with a random sample — and the high-risk programs not yet covered.
National dashboardPrograms scored and high risk by week, red flags by OIG factor, flags by territory and how programs are held.
Your scoring rulesMeal cap, repeat attendance, deck review line, label age, speaker cap alert and the high-risk line — each showing how many programs it flags now.
Governance

Built for commercial compliance: cited, decided by a person, unannounced.

Every flag cites its source lineEach red flag points to the request, RSVP list, sign-in sheet, invoice, nomination or agreement line it came from, next to the OIG factor and the rule it breaks.
A person decides every high-risk programAny program at or above the high-risk line waits for a monitor: clear, change, hold, audit or refer. A clearance carries the monitor’s note, which stays on the record.
Audits stay unannouncedThe audit plan is visible only to compliance. Sales, marketing and the bureau see a monitor arrive, never a calendar entry — the setting is locked on.
Facts, not conclusionsThe monitoring report states facts and recommends actions; it never concludes intent. It stays a draft until a monitor signs it, or signs and refers it.
Rule changes are approved and re-scoredA change to the scoring rules needs the chief compliance officer’s approval, re-scores every open program and is recorded in the audit trail.
Every step on the recordEach agent step and each human decision is logged on the program’s activity trail — who, what and when. Speaker data stays in the US region.
Configuration

Your policy, not ours

The thresholds every program is scored against are settings. Each one shows how many programs it flags right now.

SettingDefaultChoose from
Meal cap per head (standard areas)$95$60–$150, in $5 steps
Flag an HCP at this many same-topic programs in 12 months2 programs2–6 programs
Deck review line250 programs in 12 months100–600, in steps of 50
No new information when the last label change is older than12 months6–36 months
Speaker cap alert75 % of the $50,000 annual cap50–95 %
High-risk line — a person must decideScore 60Score 40–80
Random audit sample2 % of each week’s remaining programs1–10 %
Flag restaurants · wine bars, country clubs and entertainment venuesOn · on, weighted higherOn · off
Connections

Works with the systems your programs already run on

Your events platformprogram requests, RSVP lists and changes
Scanned sign-in sheetshandwritten names, NPIs, signatures and notes
HCP masterNPI, specialty, city and attendance history
Contract systemspeaker agreements, fees and the fair market value rate card
Deck library and FDA label historyapproved decks and label changes
CRM territory dataterritory plans and nomination context
What it changes

The difference, in numbers.

Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.

target
100%
of programs scored on all eight red flags
scored; about 3% seen live today
target
12of 18
unannounced audits a month aimed by the risk score
12 risk-based
6 random · no advance notice
target
9programs
a month fixed before they ran: venue or format changed, or held

“demo” = seen in the working solution, on its sample programs · “target” = the design goal, measured in the live solution · “estimated” = our estimate, including today’s live-audit share · “typical” = published source: OIG Special Fraud Alert on speaker programs (2020). Also referenced: Anti-Kickback Statute, 42 USC 1320a-7b(b) · PhRMA Code (2022). People, speakers and brands named on this page are characters in the working solution.

Questions

What commercial compliance teams ask us.

What is speaker program compliance monitoring?

Checking that each promotional speaker program is a genuine educational event and not a kickback — the venue, the meal, the topic, who attends and how the speaker was chosen and paid — before it runs and again afterwards from the sign-in sheet and invoice. Speaker Program Sentinel does this for every program and shows the evidence for each flag.

Which red flags does it score?

The eight in the OIG Special Fraud Alert on speaker programs: a restaurant or entertainment venue, no new medical information, alcohol or a meal above modest value, many programs on one topic, repeat attendees, attendees with no business reason, sales-driven speaker choice, and pay near or over the cap. Each is set against the PhRMA Code or your company rule, and the result is a 0–100 score with reasons.

Does it check programs before they run?

Yes. Each request is scored from the plan as soon as it arrives, so a monitor can change the venue or format, put the program on hold, plan an unannounced audit or refer it before the event. After the event it is scored again from the sign-in sheet and invoice.

How does it find repeat attendees?

The Repeat Attendee Tracker matches every invitee and signer to the HCP master by NPI, then name and city, and counts same-topic programs over 12 months. It also shows speakers who attend each other’s programs. A match below 0.9 goes to a person.

Will the field know when an audit is coming?

No. Audits are planned by risk and topped up with a random sample, and the audit plan is visible only to compliance. Sales, marketing and the speaker bureau never see it; the setting is locked on by policy.

Do people stay in control?

Yes. A monitor decides every program at or above the high-risk line, can confirm or dismiss each flag, and signs every monitoring report. The agents recommend actions and never conclude intent; referrals go to a named investigator with the evidence pack.

Can we set our own meal cap and thresholds?

Yes. The meal cap per head, the repeat-attendance count, the deck review line, the label age for “no new information”, the speaker cap alert, the high-risk line and the random sample are all settings. A change re-scores every open program, needs the chief compliance officer’s approval and is recorded in the audit trail.

How long does it take to go live?

The Agentic Solution Engine builds and deploys it from your requirements — your speaker program policy, meal caps, report template and a sample of past programs — and it goes live once every quality gate has passed. We will walk you through it on your own programs first.

Compliance & Ethics

More in this division.

All 5 solutions →
Payment Recipients

Every payment to a clinician or teaching hospital matched to the right registry record, with the evidence behind each match.

1-min filmThird-Party Due Diligence

Every distributor, agent and vendor checked for bribery risk before the contract is signed.

1-min filmInvestigation Builder

Every investigation starts with its evidence gathered, ordered and cited, and closes in a median of 18 days.

1-min film

See it on
your programs.

We’ll run Speaker Program Sentinel on a sample of your own past speaker programs.

Book a demoBrowse solutions