Main site

Strategy AI agent · Investor relations

Investor Relations Earnings AI Agent

Timestamp each investor statement against the audience that received it, watch for anything said to one room and not the market, and hold the corrective for a named disclosure officer.

Typical delivery
4–6 weeks
Deployment
Your stack
No signature
One clock
After launch
Agent Care

What this agent does

Assembles the statement, never the clearance

In

  1. 01

    A statement is drafted, and 17 CFR Part 243 governs who may hear it first.

  2. 02

    A remark is intentional, and § 243.100(a)(1) wants the market told simultaneously.

Reason

  1. 03

    A slip is unintentional, and § 243.100(a)(2) wants public disclosure promptly.

  2. 04

    A senior official learns of that slip, and § 243.101 starts twenty-four hours.

  3. 05

    A next NYSE open falls later, and § 243.101 makes that opening the deadline.

Decide

  1. 06

    A listed recipient hears it, and § 243.100(b)(1)(i) to (iv) names the classes.

  2. 07

    A results release goes out, and Form 8-K Item 2.02 leaves four business days.

Out

  1. 08

    A non-GAAP measure appears, and Item 10(e)(1)(i) wants the reconciliation beside it.

  2. 09

    Execute write actions only inside the approval boundaries agreed during implementation.

Product statement

Both timings belong to the agent. Nothing under Regulation FD is signed, certified or attested; the disclosure officer who clears a statement is named by your own policy, not by the rule.

Example workflow

One statement, evidence to clearance

  1. 1Investor statement receivedDraft releases, call scripts, conference notes or analyst meeting logs
  2. 2Disclosure context assembledThe remark, the audience that heard it, the rule it answers to and the hour the clock started
  3. 3Audience evidence draftedThe remark, the recipients under it, the gaps and completeness
  4. 4Controls appliedRecipient checks, simultaneity checks, clock checks and completeness confidence

No human action required

Stages 1 to 4 run unaided, and nothing is cleared at any of them — the agent is assembling, and disclosure counsel enters at the completeness gate.

5DecisionSplits at the completeness gate
Evidence sufficient

Passes to the disclosure officer to clear.

Anything thin

Adds a disclosure counsel read first.

Disclosure counsel review

The remark is held with its audience, its gaps and the hour it went out.

Clear · Append evidence · Send for counsel read

Cleared — by the disclosure officer

  1. 6Statement and transcript records updatedOnly where write access and records policy allow it
  2. 7Outcome evaluatedAudience completeness, clock coverage, counsel corrections and what review found

Corrections

Each correction from review is counted in the evaluation.

What should not run autonomously

Human approval stays in control

Outside the boundary — human approval required

8 items

  • Deciding a fact is material and non-public.
    • Judging whether a selective remark was intentional.
    • Clearing any statement for release to the market.
    • Settling when the promptness clock has expired.

  • Automation boundary

    Agent acts unaided

    • ✓Hold each statement beside the audience that actually received it.
    • ✓Start the clock at the hour a senior official spoke.
    • ✓Flag a remark that reached one room and never reached the market.
    • ✓Draft the corrective statement for its named signer.

    Nothing is cleared and nothing goes out to the market unless a named person says so.

    • Judging whether Regulation FD was breached.
    • Telling an analyst that a remark was harmless.
    • Signing the Form 8-K that furnishes the results.
  • Changes to a transcript, a script or the record.

Example output

One investor statement, annotated

Our performance reporting agent drafts campaign reporting for an agency client; this page is about what the company may say to the market, and when.

Investor statement · single remark

Illustrative example
Statement
Corridor remark, § 243.100(a)(2)
Recorded as
Drawn from the notes each attendee kept
Audience
Margin remark
Evidence of record
Conference notes, 3 August 2026
Confidence
Held uncleared
Held for
The disclosure officer, by name

As received

Built out of the conference notes and the call recording — nothing beyond those two is claimed.

What the record holds

  • Conference notes
  • Analyst email
  • Call recording

Why no signature appears here

Whether the remark was material is a judgement Regulation FD leaves to people.

What the score decides

Below the configured threshold a statement picks up a counsel read before clearance.

Action

ClearAppend evidenceSend for counsel read

Value

Where AI adds value

The same four claims, placed at the point in the workflow where each one applies.

Where the value lands

Value 01 – 04

Every statement

From the audience it reached

03Evidence

Where the evidence is used

The § 302 and § 906 certifications belong to the principal executive and principal financial officers under Rule 13a-14, which permits no power of attorney; that is Finance work, claimed nowhere here.

01Approved path

The clock starts on a slip

Under § 243.101 promptness runs no later than twenty-four hours after a senior official learns, or the commencement of the next day's trading on the New York Stock Exchange, whichever is later.

02Human review

What was checked, and not found

The Form 8-K signature block was not obtained verbatim, so none of it is quoted; Rule 12b-11(d) confirms only that a duly authorized officer signs for the registrant, which attests authority rather than the accuracy of the figures. No penalty figure is published, because none was confirmed in primary text, and no amendment to Part 243 after 85 FR 33360 was located.

04

Build an evidence trail

The statement, the audience it reached and the officer who cleared it stay on the record.

Integrations

Typical integrations

Five system groups connect to the same agent. Which of them are in scope is decided in discovery.

  • Investor relations desk

    • Meeting logs
    • analyst diary
    • Draft statements and briefing scripts

  • Finance and reporting

    • Close calendar
    • results pack
    • Results releases and reconciliations

  • Communications and web

    • Newsroom
    • investor website
    • Posted releases and webcast records

Agent

Investor statement evidence

  • Reads the audiences
  • Assembles the record
  • Holds for clearance

  • Transcripts and market data

    • Call transcripts
    • wire feeds
    • Timestamps and distribution logs

  • Observability & evaluation

    • OpenTelemetry
    • Langfuse
    • Supported monitoring/evaluation sources

Integration availability depends on the client's existing systems and API access.

Agent controls

Six drums between the model and the officer

Six drums turning in sequence, the last the slowest. What survives the last is named in the map below.

L6 · Outermost — last line of defence

Inward → L1 · closest to the model

L6Rollback / safe modeNarrow the agent to statement assembly when evaluation or production signals degrade.Roll back
L5Version monitoringTrack model, prompt and disclosure rules; the semiannual Form 10-S proposal, Release 33-11414, closed comments on 6 July 2026 and has no adopting release, so Rule 13a-13 still governs.Track
L4TraceabilityRecord each remark, the audience under it, the hour it went out and every read of the file.Record
L3Clearance gateHold the statement for the named disclosure officer; Part 243 requires no signature of any kind, so that hold is policy rather than law.Gate
L2Audience guardrailsTest each recipient against the classes at § 243.100(b)(1)(i) to (iv) and the carve-outs at paragraph (b)(2), which carry no rating-agency exception.Restrict
L1Confidence thresholdsRoute a thin statement to a counsel read; § 243.102 provides that an FD failure is not by itself a Rule 10b-5 violation.Require review
Model coreStatement assembled — the remark, its audience, the clock and completeness
L1 – L2
Test whether a statement may stand
L3
Hands the clearance to a named person
L4 – L5
Keep the statement and the audience behind it
L6
Holds the statement uncleared when signals degrade

How Nestack evaluates it

Evaluate the whole assembly — not only the statement record that comes out.

Coverage runs the whole depth of the workflow, and every layer is cut by slice.

Surface — the statement the market reads

  1. E1Final-output evaluationDid the record show which audience actually heard it?
  2. E2Step-level evaluationDid the agent read the right transcript, the right channel and the live calendar?
  3. E3Tool evaluationDid it read and write the correct statement record and the correct event?
  4. E4Confidence calibrationDo low-confidence statements actually attract more counsel corrections?
  5. E5Slice evaluationHow does performance change across specific audience types?
  6. E6Business outcomeHow many statements needed a correction before clearance?

Floor — what the market was told

Failure modes

Where each failure originates in the agent

Seven failure modes, each placed at the stage where it first bites.

Agent lifecycle

Direction of processing →

  1. 01 · Retrieval

    1 mode

    • MH-03

      Stale audience read

      The distribution list is not the one now in use.

    Stage gathers

    The events, the rooms, the remarks and the times

  2. 02 · Reasoning

    2 modes

    • MH-04

      Remark timed, not evidenced

      A statement is timed with nothing behind it.

    • MH-06

      Proposal read as adopted

      A proposed rule is worked as though final.

    Stage proposes

    The statements, their audiences and completeness

  3. 03 · Tool / write

    2 modes

    • MH-02

      Thin statement passed forward

      A statement passes on with no counsel read.

    • MH-05

      Wrong audience on the record

      The remark is bound to the wrong room.

    Stage writes

    Only where write access and approval policy allow it

  4. 04 · Output

    1 mode

    • MH-01

      Cleared, audience unrecorded

      The record shows clearance but not who heard it first.

    Stage returns

    The statement an officer clears and the market reads

  5. 05 · Change / Version

    1 mode

    • MH-07

      Silent clock drift

      The deadline shifts in a release, and the held record keeps the older one.

    Stage tracks

    Model, prompt, timing rules and audience fields

  • Sev-1 · a statement cleared on no record
  • Sev-2 · wrong audience reaches the record
  • Sev-3 · source degrades, statement held
See the complete Strategy failure-mode catalogue

Affected slices

One-to-one meetings absorb the corrections

An audience-level simultaneity figure can read clean while one-to-one analyst meetings carry most of the rework. Nestack reports the correction rate by audience, not only in total.

Slice performance — reported separately, not only in aggregate

Illustrative example
SliceFailure rateLiftLift vs. thresholdStatus
One-to-one analyst meetings7.8%3.6×Review
Investor conferences5.6%2.6×Review
Earnings calls and webcasts3.4%1.6×Watch
Results releases1.9%0.9×Normal

Bar: correction-rate lift vs. results-release baseline · scale 0–4.0× · tick marks the 2.0× review threshold

2 of 4 slices over threshold

Evidence-linked improvement

What a corridor answer costs

A cycle shuts when the selective remark is a regression case. That suite is what the next statement cleared is measured against.

Improvement cycle · five stages

Switchback — the path turns at Improve and returns at Learn

  1. 01

    Detect

    Correction rate rises on one-to-one meetings.

  2. 02

    Diagnose

    The answer given at the edge of a conference, which the transcript never carried, is replayed until one cause is left standing.

  3. 03

    Improve

    The change ships numbered, and the statements that forced it ride with it.

  4. 04

    Verify

    Nothing releases while one touched statement case is still red.

  5. 05

    Learn

    It is retained for good, and the clearance rules are amended in that same commit.

  6. Learn → Detect

    The return edge. The next statement is measured against a suite one case longer.

Typical build scope

Twelve workstreams across six weeks

The build scope read against the delivery timeline. Week structure follows the six-week plan — discovery, sources, statement assembly, evaluation, integration, then production validation and handover.

Workstream

Week 1Week 2Week 3Week 4Week 5Week 6

  1. 01Disclosure-boundary and audience-mapping work.Week 1
  2. 02Transcript, wire and IR-desk sources.Week 1, Week 2
  3. 03Statement-to-audience and clock-coverage mapping.Week 1, Week 2
  4. 04Investor statement ingestion.Week 2
  5. 05Remark, audience and hour binding.Week 2, Week 3
  6. 06Completeness scoring and review routing.Week 3
  7. 07Officer clearance workflow.Week 3, Week 4
  8. 08Newsroom and transcript integration.Week 5
  9. 09Selectivity and timing cases.Week 4, Week 5
  10. 10Guardrails and clearance controls.Week 4
  11. 11Statement-trail instrumentation.Week 5, Week 6
  12. 12Deployment, documentation and Agent Care handover.Week 6

12 workstreams · 6 weeks · bar shows the weeks a workstream is active — several run in parallel

Final scope and sequence confirmed in discovery

Engagement tiers

What each tier includes

Rows are the capabilities named in each tier's scope. Higher tiers include everything below them.

Capability✓ in scope · — not at this tierPilotOne cycle, one marketProductionProduction disclosure workflowAdvancedMultiple markets / entities
Introduced at Pilot
Statement assembly to your calendar✓✓✓
Disclosure officer release✓✓✓
Disclosed-figure baseline✓✓✓
Introduced at Production
Reporting by audience—✓✓
Officer clearance workflow in your systems—✓✓
Approved write-back—✓✓
Transcript-and-filing integration—✓✓
Introduced at Advanced
Multi-market disclosure——✓
Cross-channel statement packs——✓
Large transcript estates——✓
Multi-channel timing controls——✓
Build priceFrom $5,000From $8,000Custom quote
Final build price
Confirmed after discovery based on integrations, workflow complexity, statement volume, approval controls and deployment requirements.
Separate from build
Build pricing is separate from recurring Agent Care, which covers managed monitoring, evaluations, incidents and verified improvements after launch.

What we need from you

What you bring, and what we build with it

Each input maps to a piece of build scope and a week in the delivery timeline.

  1. 01Your reporting calendar and the audiences behind each event

    Audience mapping and statement capture

    Week 1

  2. 02Representative releases, scripts and call transcripts

    Audience binding, clock logic and the statement baseline

    Week 2

  3. 03Your standing disclosure policy and the officer it names

    Audience mapping, clock binding and the automation boundary

    Week 1

  4. 04Access to relevant APIs, feeds or exports

    Transcript, wire and IR-desk source assessment, then integration setup

    Week 2

  5. 05Statements you would not want replayed

    Timing cases and failure-mode testing

    Week 4

  6. 06What no investor statement may promise

    Completeness scoring, review routing, guardrails and release controls

    Week 3

  7. 07A named disclosure officer to clear what goes out

    Officer clearance workflow, then pilot and production validation

    Weeks 5–6

Nothing else is required

Deployment, documentation and Agent Care handover are ours.

Delivery timeline

Four phases across six weeks

Every width answers to the work rather than the page, so a single week ends up holding two phases.

Phase

W1

W2

W3

W4

W5

W6

  1. DiscoveryW1
  2. BuildW2 – W3
  3. EvaluateW4 – W5
  4. Pilot & LaunchW5 – W6

Week focus

W1Disclosure workflow discovery, audience mapping and the automation boundary

W2Source integration and the investor-statement baseline

W3Statement assembly, clock logic and release controls

W4Evaluation suite, selectivity cases and regression testing

W5Transcript integration, pilot statements and targeted corrections

W6One reporting year run under the disclosure officer, then Agent Care handover

Reading the band
Each bar runs only as far as the work it names, and the fifth stretches to take two phases.
At the end of W6
Validation closes on live statements, and Agent Care picks up the watch.
Duration
Six-week plan shown · typical delivery 4–6 weeks depending on scope confirmed in discovery.

Next step · Strategy AI agent

Build an IR disclosure agent around the remark nobody put in the transcript.

Show us one statement and the room it was made in. The clock starts on what a senior official learned and when, not on anything anyone wrote down (§ 243.101). Under § 243.100(a)(1) an intentional selective disclosure has to reach the market at the same moment.

Typical delivery
4–6 weeks
Deployment
Your earnings cycle
No signature
One clock
After launch
Agent Care available

Nestack Agents · Investor relations

AGT-CS-09 · Agent Care available after launch

Build an IR disclosure agent

Four questions so the call starts from your actual setup — then pick a time that suits you.

About you
Your statements

No obligation. We’ll map the workflow, name the automation boundary and tell you if an agent is the wrong answer.