For MDS Coordinators

The MDS workflow, owned.

Someris AI's MDS Coordinator Agent ingests EHR data, drafts every section with line-level evidence citations, runs CMS edit checks, synthesizes prescriptive CAA narratives, and surfaces NTA + SCPA opportunities, all before you open the case.

The job today

  • 5-Day windows you miss because the chart wasn't in the inbox.
  • iQIES rejections that take 3 hours to decode and re-submit.
  • CAA narratives that read like boilerplate because nobody has 30 minutes per area.
  • $200/day NTA opportunities buried in I8000 because nobody had time to scan.
  • 730-day SCPA windows closing on coding errors you already know about.

What the agent does

Persistent ARD calendar across the census

Tracks every resident's 5-Day, 14-Day, IPA, Quarterly, Annual, Discharge obligation. Pre-stages draft assessments N days before ARD.

All 21 sections drafted with evidence

A/C/D/E/GG/H/I/K/M/N via the EHR prefill orchestrator; B/F/G/J/L/M-detail/N-regimen/P/Q via specialized scanners. Every value cites a source.

Prescriptive CAA narratives

Not boilerplate. The 20 CAA decision narratives are synthesized from clinical evidence + RAI Manual chunks, with confidence floors.

iQIES rejection auto-resolver

Parses the rejection report against the 818-record CMS catalog, maps each error to specific MDS items, proposes fixes with citations.

730-day SCPA scanner

Continuous scan for new diagnoses, GG declines, missed NTA codes that warrant SCPA. Ranked by $-impact + days remaining.

Provider-query auto-transcription

Provider responds in plain text. The agent transcribes the answer into the correct MDS item with confidence-gated auto-apply.

How we earn trust

Evidence-anchored
Every clinical claim cites a source document by date + line. No invented values. Coordinators see exactly where the agent looked.
Survey-defensible
Every action is HMAC-signed and replayable. Run the survey-readiness simulator the day before. Find what a CMS surveyor would find.
You stay in control
Tiered autonomy per facility. Default: agent drafts; you approve. Earn trust before increasing autonomy.
See it on your facility.

We run the agent in shadow mode against 30 days of your real attested cases and hand you the ROI report. No code changes on your side.

Book a 30-day shadow backtest