Announcements

Prior auth is mostly a documentation problem.

Which policy applies. What the chart already supports. What’s missing before it goes to the payer. The Prior-Authorization Workflow in the AKOS Arsenal works through that in order:

Policy stack → Determination + missing inputs → Medical-necessity attestation → Payer submission

It resolves the governing LCD or NCD, cites it in the determination, flags missing documentation, and routes a review-ready packet to the payer.

Built for health systems, specialty practices, skilled nursing, and payers and MACs.

Medical Coding in the AKOS Arsenal reads the op note, extracts clinical entities, and links each one to the diagnosis it modifies. It then suggests ICD-10-CM and HCPCS codes.

Op note → Entity extraction → Relation linking → Code suggestions → Coder review

The coder stays in control. Approve, reject, or query the provider. Built for health systems, specialty practices, billing and RCM services, and payers and MACs.

This month, MEDKONG moved onto a custom Medplum app that connects to large EHR platforms like Epic and Oracle Health Millennium. One app for the four teams that touch every prior auth request, all working from the same record.

Here’s what each team gets 👇

Providers

The full EHR, plus a pre-auth check on every procedure order. Before anything is submitted, the provider sees whether prior auth is required, how the MAC is likely to respond, and exactly which documentation requirements are still unmet. One click opens the PA request, and any change to the order stays in sync with it.

Prior Authorisation Coordinators

A filterable request queue and a detail page for every request, with notes, audio, and video attachments in one place. Coordinators can edit requests, associate codes, and trace related records on a linked-records graph. Before submitting, a preview run shows Foundry’s determination, so weak requests get fixed instead of filed.

Medical Coders

A visit queue with workload filters by coder, visit type, and coding status. Each visit’s transcript sits next to its codes. AI recode suggests ICD-10 and CPT codes straight from the transcript, and an entity graph highlights the diagnoses, procedures, and body sites it pulled out, so coders can see why a code was suggested. When something’s unclear, a query thread goes straight to the provider.

MAC Reviewers

A case queue and a case page that brings together the request, supporting documents, and policy references. Review panels are editable and saved under each reviewer’s name. “Ask the machine” runs an AI check of the documentation against policy. Decision letters go out on the contractor’s letterhead, reviews carry forward between rounds, and every round produces a full read-only audit packet.

From the first order to the final decision letter, every team sees the same request, the same documents, and the same history.

Built on Palantir Foundry and AIP.

We redesigned the Dandelion Emotional Wellbeing Toolkit around the people who use it most: adults between 30 and 60, building better habits one module at a time.

The new design came from three places: feedback from the Dandelion team, their brand, and a lot of time asking what users actually need when they log in.

The Dashboard

A cleaner layout that keeps users focused on their goal. The app’s most-used features sit right there as quick actions, and a Continue Learning section previews the next module, or picks up right where they left off. A breathing activity is built into the dashboard too, for the days when that’s all you need.

The Modules

Each module now takes center stage, with next-step cards up front, so it’s always clear what comes next.

Everything else, a tap away

Every activity lives on one page, from quick exercises to the full Emotional Wellbeing Assessment.

Behind the scenes

A new backend dashboard for the Dandelion team. Admins can now see everything at a glance, from new signups and revenue to module completions and check-ins, with each period compared against the one before.

OphthoCards has a new Editorial Workflow, built for the people behind the content. Contributors draft and submit cases directly in the platform. Editors review, edit, and approve in the same place. Nothing gets copied, pasted, or lost between a doc and the live site.

Until now, getting that case online took a lot more than medicine. Owners, editors, and contributors drafted cases in Google Docs, traded edits and comments back and forth, then copied everything by hand into the website admin. Every case. Every approval.

That’s gone now.

For the ophthalmologists writing and reviewing cases, it’s hours back every week, and more of their time on the medicine instead of the formatting.

On the learner side, there’s plenty new too:

• High-yield flashcards you can swipe through
• Structured study guides
• Progress tracking to see what you’ve mastered and what needs another round
• A new dashboard with notification alerts, so nothing slips

Better tools for the people writing the cases. A sharper way to study for the people learning from them.

Purchase Journals is now available in the AKOS Arsenal, our library of pre-built tools.

The tool extracts structured data from purchase journals, turning document-based records into clean, usable data. Teams can bring that data into Foundry and AIP for reporting, reconciliation, and downstream workflows without manual data entry.

To learn more or see it in action, get in touch with the AKOS team.