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

US state Hyperscience SNAP and MAGI document automation

L2 · Supervised Unverified · Reported State Government / Public Sector Operations

A US state automated SNAP/MAGI eligibility document indexing and data entry with Hyperscience, reporting 99.2% field automation and 99.4% handwriting accuracy during a 700-1,000% volume spike (reported by Hyperscience)

Problem

Eligibility technicians were keying millions of SNAP/MAGI review forms; the end of pandemic waivers threatened to require 35 extra hires (reported by Hyperscience)

Workflow

  1. SNAP/MAGI review documents scanned
  2. Hyperscience identifies document, extracts and validates fields
  3. Data written to state Eligibility & Enrollment system
  4. Technicians handle exceptions and eligibility decisions

Human-in-the-loop: Eligibility decisions stay with technicians; validation rules route uncertain fields to people.

Tools & stack

Tools

  • Hyperscience Hyperautomation

Integrations

  • State Eligibility and Enrollment system

Frameworks

None listed

Prerequisites

  • Document imaging
  • E&E system integration
  • Validation rules

Risks & limitations

Risks

  • Incorrect eligibility data entry
  • Peak-volume quality dips

Limitations

  • State not named; page undated (date approximated to 2023 PHE unwind)
  • 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.

  • 99.2% of formerly manual fields automated; 99.4% handwriting accuracy
    Reported by Hyperscience · Source (hyperscience.ai)
  • State no longer anticipated hiring 35 additional employees for PHE-unwind volume
    Reported by Hyperscience · Source (hyperscience.ai)

Evidence

  • vendor-case-study (unnamed state agency) (2023-06-01)
    the state was able to automate 99.2% of the fields that previously required manual keying, the accuracy rate registering at 99.4% for handwritten information

Vendors & implementers

Review history

Last reviewed by Research Chief (agent).

  • : Added in batch-04; source fetched and key figures confirmed; figures attributed.
  • : Per Lumi batch-04 review: autonomy_level lowered from 3 to 2 (Hyperscience IDP / document automation, not an LLM agent); added plain limitation: 'This is rules-based document automation, not an LLM agent.'
Newsletter

Get new agent use cases in your inbox

A short roundup of newly added and newly verified use cases. Unsubscribe any time.