Ask Sage
The conversational front door. Ask a how-to, look something up, or hand Sage a task — it answers using tools gated to what you can access. One chat drawer, on every page.
Meet Sage — your agency's AI
One AI brain across your whole agency. Its name is Sage.
Sage isn't a bolt-on chatbot. It's one grounded AI woven through Carelytic — it drafts your SOAP notes, scrubs claims before they go out, reads a faxed referral and pre-fills the chart, answers "who viewed this patient last week," and builds a report from a plain-English question. Every facet is the same Sage, and it only ever sees the data you could already see.
And Sage doesn't invent your clinical data — it reads the real chart, runs the real query, and answers from the real result. Grounded and OIG-safe by design.
Included in every tier. No AI add-on, no per-seat AI fee, no per-claim AI charge.
One brain, many facets
Not five AI widgets. One Sage, everywhere the work happens.
Most vendors bolt a separate AI feature onto each module, each with its own quirks. Carelytic has a single AI engine — Sage — and every AI surface below is a facet of it. That's why they all behave the same way: grounded in your real data, gated by the same permissions as your staff, and getting smarter from the same shared learning loop.
AI Reports
Describe the report you want in plain English — "LUPA-risk visits by branch this month" — and Sage builds a live, tenant-scoped report plus a grounded narrative. Never text-to-SQL: it picks from a validated whitelist, the engine runs the real aggregate.
Audit AI Search
"Who logged in yesterday?" "Who viewed this patient's chart last week?" Sage turns the question into a permission-gated audit query and answers with the real names and rows — surveillance-grade access logs, in plain English.
AI Document Intake
Drop a referral packet, face sheet, or onboarding PDFs on the new-client or employee wizard. Sage reads them — native PDF + vision — and pre-fills the form field by field. Extract → review → confirm. Nothing saves until an admin approves each value.
AI SOAP Draft
Drafts Subjective / Objective / Assessment / Plan from the structured findings the clinician already entered — vitals, body-system flags, comments. The clinician reviews, edits, and signs. Drop-in on every schema form.
AI Pre-Sign QA
Scans a chart for CMS Conditions of Participation gaps before the clinician signs — internal contradictions, missing homebound rationale, missing med reconciliation. Fix or override, with the reason logged.
AI PDGM Coding Review OIG-safe
Flags RTP risk, ICD-10 specificity gaps, and doc-vs-code mismatches — every finding cites the chart text it came from. Surfaces downcoding gaps as well as upcoding risk. Never recommends a change purely to lift reimbursement.
Sage for Billing
Two facets in your revenue cycle: a pre-bill claim scrub that explains what would reject before you transmit, and a denial analyst that reads a denied claim and recommends appeal, correct-and-resubmit, or write-off. Advisory only — Sage never edits or submits a claim itself.
AI Carry-Forward & POC
Smart-copies the patient's prior visit — findings and narratives, never vitals or signatures. And it drafts the CMS-485 narrative sections (§21 Orders, §22 SMART Goals) from the locked OASIS, not a canned template library.
Sage learns your agency
Correct Sage once. It remembers — for the right people, after approval.
When a biller, DON, or coder catches Sage getting something wrong for your agency, they don't file a support ticket and wait for a model update. They teach Sage — a one-click correction from anywhere in the app. It becomes a proposed Lesson.
A Lesson does nothing until a Carelytic admin approves it. Once approved, Sage applies it instantly the next time the situation comes up — no retraining, no fine-tuning, no code change. And because it's retrieval rather than retraining, every Lesson is reversible and fully auditable.
How a correction becomes knowledge
- Teach. Any authorized user flags a correction from the surface where they found it.
- Propose. It's captured as a Lesson — scoped to your agency, not pushed platform-wide.
- Approve. A super admin reviews and approves before it can ever steer clinical or billing AI.
- Apply. Sage injects the approved guidance into the relevant work — instantly, everywhere it's relevant.
- Reverse. Wrong or outdated? Pause or archive the Lesson. It stops applying immediately.
A thumbs-up / thumbs-down on any Sage answer feeds the same loop. A real hard rule ("payer X always needs modifier Y") gets promoted out of the prompt and hardened into enforced configuration — where it's applied 100% of the time.
Why the guardrails matter
The OIG is watching AI in Medicare. So is Sage.
The HHS Office of Inspector General and the DOJ-HHS working group have been explicit: AI tools that recommend coding changes purely to maximize reimbursement are a fraud-enforcement priority — the line the Amedisys and Kaiser False Claims Act settlements were drawn on. Sage is built to live on the right side of that line, and to keep your agency on it too.
Evidence first
Every finding cites the chart text it came from. The clinician sees what Sage saw — no opaque "score increased."
Bidirectional flags
Coding Review surfaces downcoding gaps as well as upcoding risk. Sage checks accuracy, it doesn't optimize for revenue.
Human sign-off
Sage never auto-applies a coding change, never edits a claim, never submits. A person reviews every recommendation and signs.
Grounded, not guessing
For anything touching patient data, Sage picks from a validated whitelist and the app runs the real query. Data comes from your database, never from the model's imagination.
Same gates as your staff
Every Sage query runs tenant-scoped and permission-gated through the exact code path your staff use. Sage can't reach data the person asking couldn't already reach.
An outage is never silent
If Sage is ever unavailable, the workflow falls back to the manual path and your admins are alerted — no lost work, no mystery blank screen.
Built to stay accurate on its own
New field today, Sage sees it tomorrow — with no code change.
The hard part of clinical AI isn't the first demo — it's staying correct as your agency's data grows. Sage is engineered so it maintains itself.
Auto-discovery — no hand-maintained field lists
What Sage can see is derived from the data models at runtime. Add a new field to a chart or a report, and it flows to Sage automatically — no one has to remember to "tell the AI about it." This is the single biggest reason Sage doesn't drift out of date.
Your PHI stays governed
Sage runs on a frontier LLM (Anthropic's Claude), right-sized per task to stay fast and affordable. It's covered by the same HIPAA-aware architecture, RBAC, and audit trail as the rest of Carelytic — every AI run is logged and reviewable.