GEICO OS

What are we working on?

Build governed apps, documents, and workflows with your organization's context.

Starter workspaces

Claims Resolution CenterConnect first notice of loss, coverage, evidence, repair, rental, payment, and investigation workflows.

Live App

Policy & Claims Platform BriefPlan resilient services across quote, policy, billing, digital servicing, claims, and partner handoffs.

Document

Responsible AI Portfolio ReviewEvaluate claims, service, and fraud-assistance opportunities with quality, fairness, access, and cost controls.

Slide Deck

Illustrative prototype. Uses public company context and simulated data; no connection to GEICO systems.

Build a claims resolution center that joins first notice of loss, coverage review, evidence, repair and rental coordination, payments, communications, and investigation referrals.

AI

I start with no access. I'll request only the evidence needed for this workspace.

Gatekeeper: Claims Operations
Read-only access to simulated claims, evidence status, appointments, and payment milestones. Sensitive policyholder fields remain masked.

Approve read-only access.

AI

Executed deterministic query

const claims = await env.CLAIMS.listOpen({ priority: "today" });
const evidence = await env.CLAIM_EVIDENCE.getCompleteness({ masked: true });
const referrals = await env.INVESTIGATIONS.listReferrals({ status: "review" });
app.render(ClaimsResolution, { claims, evidence, referrals });

The resolution view is ready. Coverage decisions, payments, referrals, and customer messages require human approval.

Create a platform modernization brief for quote to bind, policy and billing service, digital documents, first notice of loss, claims resolution, and partner exchange.

AI

I'll use curated company context and request only the resources needed for this workspace.

Gatekeeper: Service Portfolio + Architecture
Read-only access to simulated service ownership, dependencies, incidents, and portfolio metadata. No policyholder records are included.

AI

The brief maps customer jobs to resilient services, typed capabilities, and measurable review gates.

Create an AI portfolio review for claims triage support, policy service assistance, and fraud evidence summarization. Include task-specific evaluations and human decision gates.

AI

I'll use curated company context and request only the resources needed for this workspace.

Gatekeeper: Model Registry + AI Gateway
Read-only access to simulated model cards, evaluation summaries, aggregate spend, and masked test datasets. No prompts or policyholder data are exposed.

AI

Executed deterministic query

const portfolio = await env.MODEL_REGISTRY.list({
  domain: "insurance", include: ["owner", "evals", "risk-tier"]
});
const spend = await env.AI_GATEWAY.aggregate({ by: "workspace" });
deck.render(PortfolioReview, { portfolio, spend });

The review is ready. AI can assist with evidence and drafting; accountable people retain every coverage, payment, and investigation decision.

Claims Resolution Center

Illustrative data. This view does not report GEICO claim volumes, service levels, customers, or internal systems.

92%

Evidence complete

18

Items for review

7

Partner handoffs

4

Human approvals

Attention queue

Review: New evidence requires a coverage specialist before the next action is selected.

Monitor: A suspicious-pattern referral needs an investigator-owned evidence package.

Illustrative workflow readiness

First notice and coverage review

92%

Healthy

Repair and rental coordination

84%

Monitor

Payment and investigation readiness

69%

Review

Policy follows the data. Sharing this app re-checks the viewer against every observed resource.

Illustrative policy trace

Step	Enforced policy	Status
Request resource	Starts with no access; request is limited to Claims Operations	Approved read-only
Run deterministic query	Typed capability; credential remains isolated	Logged
Render app	Observed resources stay attached to the output	Bound
Share	Viewer permissions are checked at open time	Human controlled

Policy & Claims Platform Brief

Illustrative planning artifact. Derived from GEICO's public customer journeys, not internal architecture or plans.

GEICO - Policy & Claims Platform Brief

90-day planning view | Draft

Purpose

Improve the continuity of quote, policy service, billing, claims, and partner interactions while keeping customer identity, evidence, and decisions permission-scoped.

Jobs to be done

Priority	Journey moment	Required review
Unified service identity	Quote, bind, and manage policy	Identity + privacy review
Claims event architecture	Report, document, and resolve	Architecture + claims review
Governed partner exchange	Repair, rental, and payment	Security + service review

Operating principles

Start with a measurable job to be done, not a new tool.
Use curated company context before model knowledge.
The human owner remains accountable for every output.
An agent never receives more permission than the person using it.

Gatekeepers hold credentials, scope every resource, and preserve the observation trail when work is shared.

Delivery sequence

Days 1-30: Map customer jobs, service owners, critical dependencies, and evidence boundaries.

Days 31-60: Pilot typed Gatekeeper capabilities for one read-only service workflow.

Days 61-90: Measure reliability, handling time, customer effort, and control performance.

Control alignment

Insurance regulation, consumer privacy, identity assurance, decision auditability, partner boundaries, and human approval remain mandatory.

Responsible AI Portfolio Review

Slide 1 of 4

Responsible AI Portfolio Review

GEICO OS | Illustrative prototype

Slide 2 of 4

Public-context opportunity areas

Use case	Stage	Human owner	Next gate
Claims triage support	Candidate	Claims operations	Validate quality and fairness
Policy service assistance	Candidate	Customer service	Test grounded answers
Fraud evidence summarization	Explore	Investigations	Confirm human decision path

Candidate portfolio only. It is not a statement of GEICO initiatives, models, or automated insurance decisions.

Slide 3 of 4

Governance scorecard

100%

Human owner

3

Task evaluations

100%

Gateway routed

4

Decision gates

Illustrative target-state controls.

Slide 4 of 4

Next operating loops

Context: curate terminology, policies, and quality criteria.

Evaluation: define task-specific quality, safety, and fairness tests.

Access: bind every data resource through a Gatekeeper.

Efficiency: use code for deterministic work and models only for judgment.

Integrations

Organization-wide connections for GEICO OS. Gatekeepers hold credentials, scope resources, and log each action.

Prototype catalog. Connections and authorization states are simulated.

Gatekeepers

1

Productivity suite

Mail, calendar, documents, spreadsheets, and files

2

Collaboration

Chat, channels, meetings, and workflow notifications

3

Project tracking

Programs, epics, issues, sprints, and delivery status

4

Knowledge base

Policies, procedures, standards, and team documentation

5

Service management

IT tickets, incidents, change requests, and asset data

6

HRIS

Employee directory, organization, benefits, and lifecycle workflows

7

ERP & procurement

Finance, planning, purchasing, supply chain, and billing

8

CRM

Customer, account, partner, and service relationship data

9

Code platform

Repositories, reviews, issues, and engineering standards

10

Data platform

Governed warehouse, catalog, analytics, and reporting

11

Security operations

Alerts, cases, exposure, audit, and control evidence

12

Business intelligence

Dashboards, semantic models, and executive reporting

MCP Server Portals

Illustrative remote services available to authorized workspaces.

Security Operations

https://security.mcp.demo.example/mcp

Auto

Enterprise Data Catalog

https://data.mcp.demo.example/mcp

Needs auth

Finance & Procurement

https://finance.mcp.demo.example/mcp

Auto

People Directory

https://people.mcp.demo.example/mcp

Needs auth

Cloudflare API

https://cloudflare.mcp.demo.example/mcp

Auto

Organization Context

Shared, curated knowledge that grounds every GEICO OS workspace. Context is versioned and read-only to agents.

Public operating context: geico.com · Internal-looking documents below are illustrative.

md

company-strategy.md

Mission, operating model, annual priorities, and outcome definitions

md

brand-and-communications.md

Terminology, voice, accessibility, and approved communication patterns

md

security-standards.md

Identity, data protection, secure development, and incident requirements

md

responsible-ai-standard.md

AI risk tiers, evaluations, human oversight, and acceptable use

md

data-classification.md

Data categories, handling rules, retention, and sharing restrictions

md

architecture-principles.md

Technology standards, decision records, review criteria, and ownership

md

vendor-risk.md

Due diligence, contract controls, monitoring, and exit requirements

md

customer-experience.md

Journey definitions, service standards, and quality measures

md

operations-playbook.md

Service ownership, runbooks, escalation, continuity, and recovery

md

finance-controls.md

Planning, purchasing, expense, audit, and reporting procedures

md

people-policies.md

Hiring, onboarding, performance, leave, and workplace guidance

md

legal-and-compliance.md

Review paths, records, privacy, accessibility, and regulatory obligations

Skills

Reusable workflows for every function. The human requester owns the result.

Name	Description	Group	Source
meeting-prep	Build an agenda and briefing from authorized calendar, CRM, and document context	General	Shared library
weekly-operating-review	Create a cross-functional summary with decisions, owners, and open risks	General	Shared library
incident-response	Assemble evidence, draft updates, and preserve human approval for containment	Security	Shared library
vendor-risk-review	Compare due-diligence evidence with security and privacy standards	Security	Shared library
control-evidence-pack	Map authorized evidence to control requirements and identify gaps	Security	Shared library
architecture-review	Review a proposal against architecture principles and decision criteria	IT & Architecture	Shared library
change-impact	Map dependencies, affected services, stakeholders, and rollback requirements	IT & Architecture	Shared library
service-health-review	Summarize service levels, incidents, changes, and capacity risks	Operations	Shared library
runbook-builder	Turn a procedure into a deterministic workflow with approval gates	Operations	Shared library
ai-model-review	Summarize ownership, evaluations, drift, risk tier, and release readiness	Data & AI	Shared library
data-quality-report	Assess freshness, completeness, lineage, and policy compliance	Data & AI	Shared library
budget-variance	Compare actuals with plan and draft a finance-reviewed variance narrative	Finance	Shared library
procurement-brief	Summarize requirements, alternatives, risk, and approval status	Finance	Shared library
job-description	Draft an accessible role description from approved job architecture	HR	Shared library
onboarding-plan	Create a role-based onboarding plan without expanding system permissions	HR	Shared library
contract-intake	Extract terms, route issues, and prepare a legal review checklist	Legal	Shared library
privacy-assessment	Map a proposed workflow to data categories and privacy obligations	Legal	Shared library
account-brief	Create a customer briefing from authorized CRM and public information	Sales	Shared library
proposal-draft	Build a first draft using approved claims, pricing, and brand context	Sales	Shared library
executive-update	Turn project evidence into a concise decision-oriented update	General	Shared library

Profile

Illustrative account information for this public prototype.

Account

AI Gateway

Illustrative demo data. Visibility and controls across every AI provider GEICO uses — one console.

Requests

128,400

▲ 11% vs last mo

Tokens

342M

▲ 8% vs last mo

Est. spend

$9,120

76% of budget

Cache-hit

27%

▲ saves ~$2.4k

Error rate

0.6%

▼ 0.2 pts

p50 latency

480 ms

across providers

Models in Use

This month

Model	Route	Tokens	Spend	Share	p50 latency
Llama 3.3 70B	Workers AI	156M	$2,140	42%	310 ms
Claude	via AI Gateway	98M	$3,980	24%	720 ms
GPT-4o	via AI Gateway	61M	$2,510	18%	640 ms
Workers AI embeddings (bge)	Workers AI	27M	$190	16%	40 ms

Spend vs. Budget

9 days remaining

$9,120spent of $12,000 cap

76%

On track · ~$2,880 left with 9 days

Top Users

Demo User 0142M tok $1,180

Demo User 0231M tok $960

Demo User 0328M tok $840

Demo User 0422M tok $610

Usage by Workspace / Team

342M tokens total

AI Enablement

121M tokens · $3,240

Platform Engineering

89M tokens · $2,460

Customer Experience

62M tokens · $1,510

Enterprise Operations

41M tokens · $1,020

Security & Compliance

29M tokens · $890

Governance

Guardrails enforced by Gatekeepers + AI Gateway, with resource-scoped access, audit trails, and human approval.

Per-team allowed models

Restrict which providers each workspace can call.

Monthly spend caps

Hard limits per team; agents stop before overrun.

Data & AI $4,000Platform $3,000

PII redaction

Strip sensitive fields from prompts before they leave.

Enabled

Prompt / response logging

Full request logs retained for audit & review.

Enabled · 90-day retention

Rate limits

Per-team request ceilings to protect budgets.

600 req / min|burst 1,000

Raise Data & AI cap to $6,000

Change queued by an agent — needs a human sign-off.

Requires approval

What are we working on?

Illustrative policy trace

Resource boundaries

GEICO - Policy & Claims Platform Brief

Purpose

Jobs to be done

Operating principles

Delivery sequence

Control alignment

Workspace sources

Resource boundaries

Responsible AI Portfolio Review

Public-context opportunity areas

Governance scorecard

Next operating loops

Workspace sources

Resource boundaries

Integrations

Organization Context

Skills

Profile

AI Gateway

Models in Use

Spend vs. Budget

Usage by Workspace / Team

Governance

Per-team allowed models

Monthly spend caps

PII redaction

Prompt / response logging

Rate limits

Raise Data & AI cap to $6,000

AI Gateway Explorer

Review spend cap change