CClive by Five Sigma

Five Sigma says insurers using Clive saw adjuster productivity rise 25%, errors fall 80% and claim cycle times shrink 10%, results it published with Google Cloud. Those numbers explain why Clive, launched in July 2024 as an AI claims adjuster, has moved from pilot talk to live production at Starr and the Australian travel insurer Fast Cover in 2026. It is best for insurers, MGAs, TPAs and reinsurers that want AI to do routine claim work without ripping out their claims system.

The buying process is the main point to plan for. Clive is sold through enterprise contracts with no published prices, it is trained on your standard operating procedures, and every performance figure so far comes from Five Sigma. We rate it 7.7/10 on our scale; published pricing and public user ratings would lift it, and the product’s design is not what limits the score. We have not used Clive ourselves; this review is based on Five Sigma’s published material, a Google Cloud white paper and press coverage.

Build Details Five Sigma Shared Through Google Cloud

Five Sigma has published more detail on how Clive is built than most claims AI vendors, mostly through Google Cloud and trade press.

Two things stand out for buyers. First, the results moved in Five Sigma’s favor between the two sources: 10% shorter cycle times in the Google Cloud paper, 30% in the 2025 interview. That may reflect more mature deployments, but both are vendor figures, so they belong in a pilot plan as targets to test rather than as promises.

Second, the customer list has moved up market. In 2021 Five Sigma named insurtechs such as INSHUR and Hippo; in 2026 it named Starr, a global insurance group present on six continents. A global specialty carrier choosing Clive signals that the product can live inside complex, multi-line claims operations.

Claims Agents That Run on Five Sigma or Your Own System

Clive is a set of AI agents that work on claims. It can run inside Five Sigma’s own SaaS claims management platform or on top of another claims system, which Five Sigma stresses so that insurers can keep legacy platforms in place.

The Claim Workflow

1. Training: Clive is configured with the insurer’s standard operating procedures for each line of business, on top of Five Sigma’s own claims-handling data.
2. Intake: First notice of loss comes in faster, and Clive reads the documents attached to the claim.
3. Analysis: It analyzes the claim, flags discrepancies that could cause leakage and checks coverage.
4. Recommendation: It suggests next steps, drafts personalized messages and summarizes calls and interactions.
5. Decision: A human adjuster decides, unless the insurer has turned on straight-through processing for that claim type.

Specialist Agents

Five Sigma describes Clive as a multi-agent system. Named agents include Clive Coverage for coverage decisions and Clive Inspection for quality assurance. Insurers can start with one agent and add more. Five Sigma’s Google Cloud paper says it is also testing a controllable orchestration framework for workflows that use several agents together.

Three Levels of Automation

• Recommendations: Clive suggests, the adjuster acts.
• Alongside users: Clive handles parts of the file while adjusters manage the rest.
• Straight-through processing: Clive completes suitable claims end to end.

Managers also get portfolio-level views with KPIs and alerts on bottlenecks, at-risk claims and compliance gaps, according to Five Sigma.

Gemini on Vertex AI With LangChain Orchestration

The Google Cloud paper describes Gemini models on Vertex AI with support for text, images, voice and video, LangChain for orchestration and a chat interface for claims leaders, adjusters and policyholders, and LangSmith for evaluation. Custom evaluators check summaries for factual accuracy, complete reasoning and hallucinations, and user feedback is attached to traces to improve the system.

Adjusters Still Decide in Recommendation and Assist Modes

In recommendation and assist modes, an adjuster makes the call on every claim. Fast Cover’s announcement stressed that its claims decisions remain overseen by experienced claims professionals. Someone also needs to own the SOP configuration, decide which claim types qualify for straight-through processing and review the portfolio alerts Clive raises.

Enterprise Contracts Shaped by Deployment Choices

Five Sigma sells Clive through enterprise contracts. The main choices that shape the quote are listed below, based on Five Sigma’s published descriptions of how Clive is deployed.

Sizing the Value Clive Must Deliver to Earn Its Contract

Without published prices, the useful exercise is to size the value Clive would need to deliver to justify a contract. These scenarios apply Five Sigma’s own reported percentages to volumes and team sizes we chose for illustration.

• A 20-adjuster TPA: a 25% productivity gain is the output of 25 adjusters from a team of 20, or five extra adjusters’ worth of capacity. The contract plus integration needs to cost less than hiring and training those five people.
• An insurer with a 40-day average cycle: a 10% reduction brings it to 36 days; the later 30% figure would bring it to 28 days. Shorter cycles mean less loss adjustment expense and faster payments to policyholders.
• A carrier with 50 errors per 1,000 files: an 80% reduction leaves 10, and the 70% figure from 2025 leaves 15. If each error costs rework or leakage, multiply the errors avoided by that cost to set a ceiling on what Clive is worth.

Other costs to plan for: integration work with your claims system, internal time to document SOPs for each line of business, adjuster training, and any move onto Five Sigma’s platform if you choose the full SaaS route. Ask whether the quote covers each added agent separately, since expanding from one agent to several is the path Five Sigma describes.

Clive Compared With Agentech, Bevaya and Sixfold

Insurance AI agents split by where they sit in the policy life cycle. Clive and Agentech focus on claims, Bevaya covers claims, underwriting and servicing documents, and Sixfold works on underwriting.

Choose Clive if you want an agent that plans the claim and can close simple ones on its own. Agentech suits teams that want task automation while adjusters keep every decision. Bevaya is the better match if most of your pain is document intake across departments. More options are on our AI agents for insurance page, and enterprise buyers can compare others in AI agents for enterprise.

Carriers on Legacy Claims Systems and Specialty Insurers

• Carriers with legacy claims systems: Clive runs on top of the existing system, so you can add AI without a core replacement project.
• Specialty and commercial insurers: Starr’s global commercial and specialty claims rollout shows Clive working in complex, multi-line portfolios.
• TPAs under volume pressure: Five Sigma says teams handle higher claim volumes without adding staff.
• MGAs and travel insurers outside the US: the Fast Cover deployment included ABA payment rails and Lloyd’s bordereaux reporting.
• Claims leaders who want gradual automation: the three modes let you start with recommendations and move to straight-through processing later.

Agencies Wanting Phone Bots and Underwriting Teams

• Insurance agencies looking for a phone agent: Clive works on claims files, not inbound service calls. A voice agent such as Sonant AI fits that job.
• Underwriting teams: Clive is built for claims; underwriting tools such as Sixfold cover submissions and risk analysis.
• Buyers who need published prices: every deployment is quoted.
• Small teams without SOP documentation: Clive learns from written procedures, so teams without them face extra work before seeing results.
• Buyers who want crowd-sourced reviews before signing: Clive is sold to a small number of large claims organizations, so the evidence base is case studies and press releases rather than review sites. Plan on reference calls instead.

Vendor Percentages That Differ Between Sources

• Vendor-reported metrics: the 25%, 30%, 70% and 80% figures all come from Five Sigma. Ask for reference calls with insurers in your lines.
• Differing figures: different sources quote different cycle-time gains (10% and 30%), so agree on baseline and measurement before a pilot.
• Straight-through processing rules: decide in writing which claims Clive may close alone and how humans audit those decisions.
• Integration depth: Five Sigma says integration depth is set by the insurer, which also means the integration work and its timeline sit partly with you.
• Data and models: Clive runs on Google Gemini through Vertex AI, so check how claims data is processed and stored with your security and compliance teams.

Fast Cover Goes Live in Australian Travel Insurance

• April 20, 2026: Fast Cover put Five Sigma’s claims platform and Clive into live production, Five Sigma’s first production deployment in Australian travel insurance.
• January 26, 2026: Starr selected Five Sigma’s AI-native claims platform and Clive for the next phase of its digital claims transformation.
• November 3, 2025: In an interview with FinTech Global, Five Sigma described Clive as a multi-agent claims expert with three automation modes and reported 30% faster cycle times.

An AI Adjuster Without Ripping Out Your Claims System

For insurers and TPAs that want an AI adjuster without replacing their claims system, Clive is one of the more complete options, and the Starr and Fast Cover deployments show it working in production. It scores 7.7/10 on our scale. The product scores well on autonomy and integration approach; published pricing, a lighter setup than SOP configuration and public user ratings would lift the overall number.

The practical route is a narrow pilot: pick one line of business, start with a single agent in recommendation mode, and measure cycle time, error rate and adjuster throughput against a baseline. If you see results near Five Sigma’s reported 25% productivity gain, expanding to more agents and straight-through processing becomes an easy case to make.

Contract Cost, Adjuster Roles and Supported Claims Systems

How Much Does Clive by Five Sigma Cost?

Five Sigma sells Clive through enterprise contracts quoted per insurer. The price depends on whether Clive runs on your existing claims system or on Five Sigma’s platform, and on how many agents you switch on. You can start with a single agent.

Does Clive Replace Human Claims Adjusters?

It is designed to work with them. Insurers choose recommendations only, working alongside adjusters, or straight-through processing for suitable claims. Fast Cover, for example, keeps claims decisions overseen by experienced claims professionals.

Which Claims Systems Does Clive Work With?

Five Sigma says Clive runs on top of any claims management system through REST APIs, as well as inside its own SaaS claims platform. The insurer decides how deep the integration goes.

Which AI Models Power Clive?

Clive is built on Google’s Gemini models on Vertex AI, with LangChain for orchestration and LangSmith for evaluation, according to a white paper published by Google Cloud. The evaluators check summaries for factual accuracy and hallucinations.

How Is Clive Different From Agentech?

Both focus on claims. Clive can plan claims and, if the insurer allows, process suitable ones straight through. Agentech runs hundreds of task agents per claim file for work such as document review and quality checks, and says its agents do not make claim decisions.