CAIBots Agent Library — Enterprise Edition
Every workflow.
Every governance layer.
One AI workforce.
46 purpose-built execution agents across Financial Services, Capital Markets, and Healthcare — each shipping with a five-layer execution trace, pre-configured governance matrix, and system-of-record integration map. Deploy in days, not months.
🧠 Horizontal Platform
🏦 Financial Services
📈 Capital Markets
🏥 Healthcare
⚡ Revenue & Growth
46
Production
Agents
Agents
5
Execution
Layers
Layers
<50ms
Global
Inference
Inference
SR 11-7
Model Risk
Compliance
Compliance
90
Days to
Production
Production
24/7
Autonomous
Execution
Execution
Five-Layer Architecture
Built on a governed execution foundation.
Every CAIBots agent runs on the same enterprise stack — ingestion, cognition, governance, action, and audit — with domain intelligence adapters per vertical. Full architecture →
Layer 01–02
Horizontal Platform Core
Reusable AI capabilities — orchestration, RAG retrieval, model governance, identity, and observability. The foundation every vertical agent builds on.
6 platform agents · PLT-01–06
Layer 03–04
Domain Intelligence Agents
Vertical-specific execution agents with pre-configured governance matrices for Financial Services, Capital Markets, and Healthcare regulatory environments.
40 domain agents · 25 live
Layer 05
Governance & Audit Layer
Every inference is logged with prompt, output, model version, confidence score, and retrieval source. Every entry is immutable and exportable, so an OCC, Fed or FINRA request is answered from the record rather than reconstructed. Regulatory descriptions pending counsel validation.
Every inference logged
Filter:
Three numbers, and what each one means. 46 agents are catalogued with a complexity tier and a reference price. 25 are live today — every one of them runnable at caibots.com/demos, not a slide. 28 are profiled in detail on this page; the rest are platform components and vertical variants listed in the pricing catalogue. The remaining 21 are on the roadmap, sequenced by year.
Live today · runnable now
Twenty-five agents are already running.
These are deployed, governed, and open in a browser right now. The catalogue below is the full 46-agent library — live agents and roadmap together.
Enterprise suites · custom-built
4 governed decision suites
Annual subscription · by institution size
CU / smallCommunityRegional+
KYC-01 · KYC/AML$30,000$60,000$140,000
CU-01 · Credit Underwriting$26,000$48,000$105,000
FD-01 · Fraud Detection$26,000$52,000$120,000
IR-01 · Investment Research$20,000$40,000$80,000
Purpose-built. Each carries a derived risk engine, a cited policy layer and a coverage benchmark — 66 scenarios and 339 differential test cases across the four. A suite is 5–7 agents at one price. Suite pilots: fixed fee $40K–$110K by suite · 50% of any Assessment credits toward the pilot.
Run them now →FinAI · open taxonomy
10 agents on the Anthropic reference set
Annual subscription · by institution size
CU / smallCommunityRegional+
FinAI · regulated tier$10,000$20,000$38,000
FinAI · advanced tier$8,000$15,000$28,000
FinAI · standard tier$5,000$10,000$18,000
Built on the anthropics/financial-services taxonomy under Apache-2.0 — not a partnership. Each is a single agent, priced below the suites above. Single-agent pilot: $15K–$35K fixed, 6–8 weeks. Take FinAI where you want one capability rather than a governed decision surface.
Open FinAI →Platform layer
Included in every licence
Not priced per agent
PLT-01 OrchestratorCore · PLT-02 RAGCoreIncluded
PLT-03 ModelRouter · PLT-04 AuditIntelIncluded
PLT-05 SecureGate · PLT-06 ObserveIQIncluded
Orchestration, retrieval, model governance, audit, identity and observability — the foundation every vertical agent runs on.
Architecture →Adjacent product lines
Separate products, separate buyers
Each sold on its own motion
$95,000FV-01 · FamilyVault SovereignFamily offices $500M–$5B+. Assessment $12,500–$25,000 credits at 50% against implementation from $225,000.
$7,500 one-timeCaiOS · PaaS + SaaSPaaS licence: one payment, perpetual, unlimited bots — one of 271 seats (founding $5,000, first 100). SaaS: we build & host, $500/agent + $100/mo.
Free · Pro $190/yrCI · Career Intelligence TerminalThe career terminal for the frontier economy — live radar, managed positions, fit levers, weekly brief. Cohorts from $8,400/yr.
Assessment fees credit forward. Pilot Planning credits against the pilot; the FamilyVault assessment credits against implementation. The advisory engagements at /assessments do not credit — they are independent by design.
How the engagements ladder →What the subscription covers
IncludedAgent licence · platform layer PLT-01–06 · shared-cloud hosting · model inference · governance overlay and audit trail · policy-pack refreshes as regulations change · business-hours support
SeparateImplementation $40K–$1M · data-foundation work (entity resolution, lineage, remediation) · custom agent engineering · third-party data feeds · premium support · single-tenant deployment · regulated-domain risk premium
NeverRegulatory approval or examiner sign-off · legal, tax or investment advice · guaranteed detection rates — no precision or recall has been established against a labelled production book · replacement of the BSA officer or the filing decision
How these prices are built. Bottom-up from what the work removes from your cost base, not a tier table. Inputs: analyst loaded cost, alert volume, and a 40% automatable share — the figure Hawk AI publish from customer case studies, not our own estimate. Priced at 25% of value created, at or below the published competitor price for each segment. Annual subscription only — implementation, data-foundation work and support tiers are separate and quoted per client. See the full line-by-line breakdown.
Roadmap agents are not priced. Four have a committed build year; the rest are on request. Pricing an agent we have not built would be a commitment with no delivery date behind it, and it would make the live prices above — which do reconcile to the financial model — look notional by association. Ask and we will scope one.
Category 01
Horizontal Platform Agents
6 Agents
Cross-industry AI core — orchestration, RAG, governance, and security agents that power every vertical deployment on the CAIBots stack.
🤖
On request
roadmap
Execution Steps
1Intent classification and sub-agent routing decision
2Shared memory and context window management
3Prompt orchestration with CoT and retrieval injection
4Policy guardrail enforcement at every handoff hop
5Graceful fallback, result aggregation, response delivery
📚
On request
roadmap
Execution Steps
1Query vectorization and multi-source index search
2Multi-source relevance ranking and context assembly
3Hallucination filter and source authority scoring
4Grounded context injected into LLM prompt chain
5Source attribution logged, retrieval audit trail filed
⚙️
On request
roadmap
Execution Steps
1Task complexity, latency, and cost ceiling assessed
2Compliance jurisdiction and data residency checked
3Optimal LLM selected (GPT, Claude, Llama, open-weight on-prem)
4Request routed; automated failover on model degradation
5Model risk score logged, performance telemetry updated
📋
On request
roadmap
Execution Steps
1Every inference logged: prompt, output, model, confidence
2Model risk scored; anomaly patterns flagged automatically
3Hallucination rate and drift detection thresholds checked
4Immutable audit log written to tamper-evident store
5SR 11-7 / OCC / FINRA export generated on demand
🔑
On request
roadmap
Execution Steps
1SAML/OIDC/OAuth 2.0 federation validated on request
2Per-client namespace isolation applied; scope verified
3RBAC enforced per agent, data source, and model class
4Token quota and API rate-limit policies enforced
5Access event logged to audit trail with user and scope
👁️
On request
roadmap
Execution Steps
1Cost-per-token metering collected per agent and client
2Inference latency p50/p95/p99 tracked in real time
3Output quality scored; anomaly patterns surface
4Budget threshold alert dispatched to platform team
5Executive dashboard and cross-client analytics updated
Category 02
Financial Services Agents
6 Agents
SR 11-7 aligned agents for AML/KYC, onboarding, claims, regulatory monitoring, and credit — grounded in live institutional data.
🛡️
On request
roadmap
Execution Steps
1Ingests ownership registry, OFAC, PEP, and adverse media feeds
2Cognition classifies risk delta and determines re-score requirement
3Governance checks compliance matrix — auto vs. escalate
4Writes risk update to Core KYC, opens Salesforce case
5SAR pre-file generated, CCO notified, full audit logged
🎯
Available
2027
Execution Steps
1Retrieves lead data from CRM, enriches with external data sources
2Cognition scores: AUM tier, product fit, geographic coverage
3Compliance check: suitability screening, accredited investor status
4Routes to advisor queue in Salesforce, creates opportunity record
5Meeting prep package assembled and shared with advisor
✅
Available
2027
Execution Steps
1Document verification: ID, proof of address, income documentation
2KYC/AML check against sanctions lists, PEP databases
3Governance: auto-approve within policy, complex cases → manual
4Core banking account creation, product assignment, CRM record
5Welcome email dispatched, documentation filed, audit complete
📋
On request
roadmap
Execution Steps
1Intake validation: policy number, coverage verification, completeness
2Fraud scoring and policy exclusion analysis applied
3Adjudication rules engine — approve, deny, or escalate
4Claims system updated, payment instruction or denial letter generated
5Customer notified, full decision rationale logged for audit
📡
On request
roadmap
Execution Steps
1Ingests regulatory feeds across FRB, OCC, FINRA, SEC, CFPB
2Cognition classifies: applicability, impact level, deadline
3Routes to relevant compliance team with impact summary
4Creates tracking record in compliance system with deadline
5Calendar events set, escalation timeline enforced automatically
⚖️
Available
2027
Execution Steps
1Portfolio batch retrieval — utilization rates and delinquency signals
2Cognition: risk-based limit adjustment recommendation per account
3Governance: auto-execute ±$5K, senior approval above threshold
4Core banking limit update, notification to customer
5Portfolio dashboard updated, change log archived for audit
Category 03
Capital Markets Agents
6 Agents
MiFID II, Dodd-Frank, and Basel III aligned agents for transaction reporting, credit underwriting, counterparty risk, and trade surveillance.
📊
Available
2028
Execution Steps
1Retrieves all T-day trades from OMS, data warehouse, Bloomberg
2Normalizes to RTS 22 schema, enriches with LEI codes
3Validates all fields — flags exceptions with specific error codes
4Submits to ARM platform — retrieves confirmation reference
5Exception emails dispatched, compliance system updated, log archived
🏛️
On request
roadmap
Execution Steps
1Retrieves financials from LOS, pulls D&B, credit bureau, industry data
2Spreads statements, calculates DSCR, leverage, coverage ratios
3Cognition generates credit memo with recommendation and covenants
4Governance routes to credit officer for approval (tier-based)
5Post-approval: offer letter, SAP facility creation, Salesforce update
🔍
On request
roadmap
Execution Steps
1Real-time exposure data ingested from risk and trading systems
2Threshold monitoring: credit limits, concentration, covenant triggers
3Cognition: breach classification, remediation recommendation
4Governance: auto-alert below threshold, risk officer above
5Risk system updated, remediation plan executed or routed
🕵️
On request
roadmap
Execution Steps
1Ingests trade, order, and communications data streams
2Cognition applies market abuse detection algorithms and patterns
3Suspicious activity classified by alert type and severity
4Alert created in compliance system, case documentation assembled
5Compliance officer notified, SLA tracking initiated
💼
On request
roadmap
Execution Steps
1Retrieves live client positions, market data, and mandates
2Portfolio analytics, rebalancing scenarios, risk metrics computed
3Suitability and fiduciary compliance check executed per client
4Client-ready proposal or report generated in advisor's voice
5Regulatory change monitoring alert filed if mandate impacted
🔔
On request
roadmap
Execution Steps
1ICT incident detected via monitoring feed or manual trigger
2Severity classified against DORA Art. 19 classification framework
3Governance: auto-report minor, escalate major incidents to team
4Regulator portal report drafted and filed within DORA timeline
5Incident log finalized, post-incident review workflow triggered
Category 04
Healthcare & Life Sciences Agents
6 Agents
HIPAA-compliant clinical intelligence, prior auth, claims processing, and patient engagement agents grounded in EHR data and clinical guidelines.
🩺
On request
roadmap
Execution Steps
1Retrieves clinical record, diagnosis, procedure from Epic EMR
2Payer eligibility check and criteria matching against payer policy
3Governance: auto-submit (criteria met) vs. physician review (complex)
4Electronic submission to payer API, confirmation retrieved
5EMR updated, physician notified, PHI access logged per HIPAA
📝
On request
roadmap
Execution Steps
1Reads clinical notes, labs, vitals from EMR in real time
2Identifies documentation gaps, conflicting diagnoses, missing codes
3Generates CDI query or inline suggestion for physician review
4Physician approves, EMR updated, DRG recalculation triggered
5Revenue impact logged, quality metric updated, audit trail filed
🏠
On request
roadmap
Execution Steps
1Clinical record reviewed: diagnosis, functional status, payer, support
2Post-acute care options ranked by clinical fit and payer coverage
3Discharge plan drafted, care team notified for approval
4Referral orders placed in EMR, transitions coordinator alerted
5Patient and family communication sent, follow-up schedule set
🔬
On request
roadmap
Execution Steps
1Retrieves patient clinical record — diagnosis, labs, treatment history
2Matches against trial registry eligibility criteria
3Governance: physician approval required before patient contact
4Coordinator notified with patient eligibility summary
5Trial management system updated, HIPAA access log filed
⚠️
On request
roadmap
Execution Steps
1Monitors EMR for adverse event signal patterns in notes, labs, vitals
2Cognition classifies: event type, severity, causality assessment
3Governance: physician review required for all adverse event decisions
4Clinical alert in EMR, care team notification, safety officer flagged
5FDA MedWatch pre-file generated if reportable threshold met
💰
On request
roadmap
Execution Steps
1Aggregates denial data by payer, code, and denial reason
2Cognition identifies systemic denial patterns and root causes
3Correction playbook generated per payer and denial category
4Rework queue populated with corrected claims for resubmission
5Revenue cycle dashboard updated with denial trend metrics
Category 05
Revenue & Growth Agents
4 Agents
Speed-to-lead, appointment setting, market intelligence, and conversation-to-close agents that build compounding revenue pipelines autonomously.
⚡
On request
roadmap
Execution Steps
1Lead form or inbound inquiry captured across all channels
2Enrichment: company data, contact verification, intent scoring
3Personalized outreach drafted and sent within 90 seconds
4CRM record created, lead assigned to rep queue
5Follow-up sequence initiated, engagement signals tracked
📅
On request
roadmap
Execution Steps
1Prospect list pulled from CRM by ICP criteria
2Personalized outreach sequence deployed across channels
3Positive response detected; calendar availability matched
4Meeting booked, calendar invite sent, CRM opportunity created
5Pre-meeting brief assembled for sales rep, reminders dispatched
📊
On request
roadmap
Execution Steps
1Multi-source ingestion: news, filings, competitor feeds, social signals
2Competitive signals classified by relevance and urgency
3Executive briefing synthesized with source attribution
4Delivered to inbox, Slack, or dashboard on defined cadence
5Natural language Q&A layer activated for follow-up queries
🤝
On request
roadmap
Execution Steps
1Health scores computed from product usage, support tickets, NPS
2Churn risk flagged; renewal timeline and value gap identified
3Personalized retention outreach drafted and queued
4CS rep briefed with account context and recommended next action
5Outcome tracked, engagement model updated continuously
Custom Agents
Don't see your workflow?
We scope it in two weeks and agree a build window with you.
Pre-built agents cover the most common regulated workflows. Custom agents are scoped, configured, and deployed to production in the same 90-day window — on the same five-layer architecture with the same governance controls.
Princeton, NJ · contact@caibots.com · +1 (609) 721-2815