Integration · Majesco
Claims automation that runs alongside Majesco Claims for P&C
Copilot summarises what is already on the claim. Hesper goes and gets what is missing, checks it, and posts the result back with sources attached.
Last updated
In short
Hesper AI connects to Majesco Claims for P&C as an ordinary API Store subscriber, so Majesco keeps the claim, the reserves, the payments and the audit trail. Hesper reads the claim, its parties, its attachments and the correspondence thread, works coverage, investigation, settlement and recovery on it, then posts a cited findings note, an evidence attachment and a task back onto the claim. Where Copilot summarises what is already on the file, Hesper goes and gets what is not. Every decision stays with the adjuster.
Does Hesper replace Majesco Claims?
No. Majesco Claims for P&C stays the system of record: the claim, the reserves, the payments, the audit trail, and the screen your adjusters live in. Hesper administers no policy, sets no reserve of its own and moves no money. It is an evidence and automation layer bolted to the side of the core, not a replacement for any part of it.
That distinction matters commercially. Replacing a core claims suite is a multi-year programme with a bureau content migration and a regression test plan attached to it. Hesper arrives as a subscription in your own API administrator, scoped to one programme, and it can be throttled or revoked there without touching a line of configuration.
It matters technically too. Majesco describes Claims for P&C as bringing first notice of loss through to closure onto one platform, with its own Copilot and task-specific agents working inside adjuster screens. That is the system recording and governing the claim: who did what, when, and under what authority. Hesper is built for the work that generates those records in the first place: reading the whole file, pulling and testing the evidence, and drafting the output a person then signs. One system governs the claim, the other works it.
How does Hesper get the claim file?
Over the interfaces Majesco already publishes. Majesco runs an API Management platform of its own, with an API Store for browsing and testing specifications, a gateway for access, and an administrator for subscriptions and policies, and it runs the EcoExchange marketplace so third-party capability can reach the core over APIs and microservices. Most carriers keep batch and file-based feeds running beside the API estate too. Hesper reads through whichever of those routes your team chooses to open, arriving as an ordinary API Store subscriber rather than anything privileged. No desktop install, no screen reading, no privileged back door.
- A trigger, typically a new attachment or a task change, since Majesco keeps two-way email and attachments on the claim and both are useful moments to look again. A nightly queue filtered by programme, line and severity works for carriers who would rather batch it.
- The claim as Majesco holds it: loss facts and parties, the coverage and limits in play, reserves and payments to date, and the correspondence thread that usually explains more than the structured fields do.
- The documents and attachments on it: ACORD forms, police and fire reports, medical records, estimates, invoices, photos, recorded-statement transcripts and correspondence, including the two-way email Majesco keeps on the claim.
- A written scope. Carriers running CoreConnect Claims behind a delegated programme usually draw it per programme rather than per line, because the programme is the boundary the binding authority and the capacity provider's audit rights are written against.
Pull, push and scheduled batch are all viable. Because Majesco already governs third-party access through subscriptions and policies in its API administrator, the first slice is usually defined there rather than in a new piece of middleware: one programme or line, one trigger, one subscription, widened once the output has been checked against files your own adjusters closed.
What does Hesper write back?
Output lands in the three places a Majesco user already checks. Majesco's own product page describes notes, attachments and task management as the furniture of the claim file, so those are the slots Hesper's output fills. The claim number, its history and its workflow stay exactly as they were.
- A findings note in plain language, where each assertion points at the attachment, the email in the thread, or the external record that supports it.
- An evidence attachment holding the source records, a reconstructed sequence of what happened when, and each place two documents disagree with both named.
- Drafts that belong to your team the moment they land: a coverage position, a reservation-of-rights or denial rationale, a response to a demand, or a subrogation demand ready to go out over an adjuster's name.
- A task naming what Hesper thinks should happen next, raised against the adjuster or the supervisor so it enters Majesco's own task management rather than an inbox.
The absences are deliberate: no payment, no reserve movement Hesper decided on, no denial, no closure. Money and coverage outcomes stay behind your own approval chain. Each piece of Hesper output enters Majesco as a draft awaiting a person, and the approval, edit or rejection is stamped on the claim the same way it would be for work an adjuster did by hand.
What changes for the adjuster?
Majesco already puts a lot in front of the adjuster: notes, attachments, two-way email and task management on one claim, with Copilot able to summarise a loss notice or draft a letter without leaving the screen. What none of that does is go and get anything. The adjuster still reads the estimate against the recorded statement, still chases the medical record that never arrived, and still has to decide whether the vendor on the invoice is real. A summary of a thin file is still thin.
Hesper fills the file before Copilot ever summarises it. The retrieval, the cross-checking and the vendor verification happen first, so what the adjuster meets is a worked claim with citations rather than a tidy restatement of what was already there. Files with nothing wrong in them clear with the checks on the record. Files with something wrong in them arrive with the inconsistency named and the two documents that disagree sitting next to each other.
No decision right moves. Coverage, reserves, settlement authority, referral and denial all stay with your adjusters and supervisors, inside the authority matrix your programme already runs, recorded in Majesco. Claim file review sets out what a completed workup contains, and claims triage covers how a claim is scored and routed before one is run.
What does a Majesco Claims integration actually involve?
Less than carriers expect, and the shape of it is the same everywhere. No go-live date belongs on a web page, because your environment sets it and not us. What moves it is the deployment, the release cadence, the length of your security and data review, and whether the first slice is one programme or five. The sequence itself does not change.
- Scope: one line or one programme, one trigger, and a written definition of the claims that are permanently out of scope.
- Access: a subscription against a non-production environment, with the field, attachment and correspondence set named in writing and the policies set in your own API administrator.
- Calibration: a batch of already-closed Majesco claims goes through Hesper and your team scores the output against the outcome the file actually reached, which is also where the field and attachment mapping gets corrected.
- Write-back: notes, attachments and tasks start appearing on live claims, in observe-only mode first wherever a capacity provider wants sight of it before it counts.
- Widen: additional triggers, additional programmes or lines, and more of the lifecycle, released at the pace your claims leadership sets.
A go-live date offered before anyone has seen the release calendar is a guess dressed as a commitment. Bring the architecture you actually run and we will scope against it in the room. Pricing is set the same way, against the book and the programmes in scope, which is why no number appears on this page.
Where does the claim data go, and who can see it?
Connections are encrypted in transit and the data is encrypted at rest. Hesper is SOC 2 Type I attested and GDPR-ready. No part of your book is used to train models, and claim data together with Hesper's output is kept for the subscription term unless you say otherwise. Where a programme's delegated authority agreement restricts where data may sit, that restriction is written into the scope rather than handled by exception.
Each agent step is logged with the sources it used, the reasoning it followed and the time it ran. Majesco's own audit trail stays the authoritative account of who did what to the claim. Hesper's log sits beneath it and accounts for the work behind every entry Hesper added, which is what a capacity provider or a regulator ends up reading side by side.
What does Hesper do, and what does your team decide?
| Hesper's agents | Your team |
|---|---|
| Reads the claim, its parties and every document and attachment on the file | Decides which programmes, lines and severity bands Hesper is allowed to work |
| Runs coverage, investigation, valuation and recovery checks against one claim simultaneously | Sets the authority matrix those results are routed under |
| Validates claimants, providers, vendors and addresses against external records | Approves which external sources may be used |
| Posts a cited findings note, an evidence attachment and editable drafts onto the claim | Edits, signs and owns every document that leaves the file |
| Raises a task on the claim naming the action it recommends next | Holds every coverage, reserve, settlement, referral and denial decision |
| Screens every file for subrogation and salvage before it closes | Makes the pursue or waive call |
| Logs every step with its sources and a timestamp | Keeps Majesco as the authoritative record |
What does the output look like?
- Status
- Illustrative walkthrough, not a customer file. No Hesper customer, pilot or prospect is described here.
- Claim
- Property, non-weather water loss, single insured dwelling
- Trigger
- A mitigation invoice lands as an attachment on an open water-damage claim
- Read from Majesco
- Loss facts, parties, coverages, payments to date, and the documents on the claim
- Agents run
- Coverage and endorsement analysis, document forensics, contractor and entity verification, timeline reconstruction
- Written back
- A findings note, an evidence attachment, and a task raised against the adjuster
Evidence cited
- The mitigation invoice's metadata shows the file was regenerated after the date printed on it, with two line items added.
- The contractor's billing address resolves to a mailbox service, and the same address appears on three other open claims in the book.
- The reported date of loss sits four days before the first photograph on the file, and the moisture readings in the report do not move across that gap.
Recommended next action: Send it to the investigation queue with the evidence attachment on it, and hold the mitigation payment until a second inspection comes back. None of that is Hesper's to do: the payment, the denial and the reserve all belong to the adjuster and the supervisor, in Majesco. This walkthrough is illustrative and does not describe a real claim or a real customer.
Frequently asked questions
Is Hesper a Majesco partner or listed on the EcoExchange marketplace?
No. Hesper is not a Majesco partner, is not listed on EcoExchange, and holds no Majesco certification. We would rather print that than let a logo on a page imply an endorsement nobody gave us. None is needed. Hesper subscribes through the same API Store and gateway any other third party uses to reach the core.
Majesco already ships Copilot and its own AI agents. Why add Hesper?
They solve different problems. Majesco describes Copilot and its role-aware agents as working inside adjuster screens to summarise loss notices and documents, triage and assign work and draft correspondence. That is assistance at the desk, on what is already on the file. Hesper does the work that fills the file: pulling records, checking entities and vendors against external sources, running document forensics, reconstructing a timeline, and producing a cited evidence package. A summary of a thin file is still thin.
Does this work with Majesco P&C CoreConnect Claims as well as the P&C Intelligent Core Suite?
Both are workable, and the difference is mostly in which interfaces are open and who controls the release calendar rather than in what Hesper needs. Majesco positions CoreConnect Claims at MGAs, MGUs and smaller carriers, which is often where a small claims team is carrying the most volume per head. See claims automation for MGAs for how that shape usually starts.
Will this force changes to our Majesco configuration?
It should not need to. Notes, attachments and tasks are already the furniture of a Majesco claim, and those are the only objects Hesper touches. Some carriers add a field or a queue purely so Hesper's output is easy to slice in their own reporting, which is a convenience rather than a dependency.
Can Hesper pay, deny or close a claim on its own?
It cannot. Payment release, reserve movement, denial and closure are all outside what Hesper is permitted to do. Its output stops at evidence, findings, drafts and a task on the claim naming what it thinks should happen next. Somebody with the authority to decide then decides, and Majesco records that person against the decision.
We already score claims with a fraud tool alongside Majesco. Where does Hesper fit?
It does not have to, and on most books it should not. Hesper carries its own fraud detection and runs standalone, but where FRISS, Shift or Verisk is already scoring claims that score is an input Hesper uses rather than duplicates. The unserved half of the problem is what follows the flag. Around 10% of P&C claims involve fraud, a manual workup runs 14 or more days, and one investigator is carrying 200 or more files, so the flag mostly sits. Hesper works it. See fraud investigation.
Other claims systems Hesper works beside
Book a demo on a Majesco claim
A short walkthrough on your own claim types and programmes: what Hesper reads out of Majesco, what it posts back onto the claim, and where your team's authority starts.
Book a demo