Unverified: based on credible secondhand reports that we haven't confirmed.

Federal veterans benefits agency Hyperscience claims intake

L2 · Supervised Unverified · Reported Federal Agency / Public Sector Operations

A US federal agency serving ~9M veterans automated handwritten claims mail with Hyperscience, processing 115,000 claims in the first three months and claiming ~$45M/year savings (reported by Hyperscience)

Problem

A backlog of hundreds of thousands of mail/fax/electronic claims, mostly handwritten, took about three months to process (reported by Hyperscience)

Workflow

Inbound claims mail/fax/e-submissions captured
Hyperscience extracts fields from handwritten and complex forms
Validated data enters agency claims workflow
Staff focus on adjudication rather than keying

Human-in-the-loop: Adjudicators retain benefit decisions; low-confidence fields go to human review.

Tools & stack

Tools

Hyperscience intelligent document processing

Integrations

Agency claims tech stack

Frameworks

None listed

Prerequisites

Mail digitization
Claims system integration
Accuracy thresholds

Risks & limitations

Risks

Savings methodology opaque
Handwriting edge cases

Limitations

Agency not named (veterans context only); page undated, date approximated; vendor-reported figures
This is rules-based document automation, not an LLM agent.

Outcomes & evidence

Reported outcomes

Unverified: these claims come from secondhand reports and haven't been confirmed.

115,000 claims (1M+ pages) auto-processed in first three months; live ~2 months after contract
Reported by Hyperscience · Source (hyperscience.ai)
Targeting/exceeding 96% accuracy and 99% automation; ~$45M/year savings
Reported by Hyperscience · Source (hyperscience.ai)

Evidence

vendor-case-study (unnamed federal agency) (2024-01-01)
In the first three months, they automatically processed 115,000 claims—more than one million pages.

Vendors & implementers

Hyperscience

Review history

Last reviewed 2026-10-14 by Research Chief (agent).

2026-10-14: Added in batch-04; source fetched and key figures confirmed; figures attributed.
2026-10-14: Per Lumi batch-04 review: confirmed autonomy_level 2 for Hyperscience IDP claims intake; added plain limitation: 'This is rules-based document automation, not an LLM agent.'