CloudKites
Emu: a calmer, more capable on-device assistant
Refinements to everyday drafting, reading, and multi-modal help — still running entirely on your own device, with no account and no telemetry.
CloudKites
ケアの現場で実現する、AIネイティブな医療画像。
オンプレミスまたはハイブリッドで導入できる信頼性の高い臨床AI。医療画像のためのAIネイティブPACSから始まり、エージェント型、そして最終的には身体性を備えた臨床アシスタントへと発展します。
臨床的信頼のための設計
私たちが作るすべてに、4つの原則が貫かれています。
オンプレミスまたはハイブリッド
ポッドはお好きな場所に導入できます — 完全オンプレミスまたはハイブリッド。
標準準拠
DICOM・HL7に対応しHIPAAを考慮、既存のシステムに適合します。
人間が関与する判断
AIが支援し下書きを作成し、臨床医が確認して署名します。
無料ツール
一般医にもすべての人にも役立つ臨床AI。
Edge and pod
Two layers: the edge at the point of care, the pod behind it
CloudKites is an AI-native PACS built in two layers. Edge software assists healthcare workers in real time where care happens; the pod is the information system and archive behind it — deployable on-prem or hybrid by your choice.
The edge — EndoEdge, PathoEdge, TomoEdge — runs on the device where the work happens. It assists the clinician in the moment and captures the study as it is made: images, video clips, whole-slide images. Because it lives at the point of care, the assistance arrives at the pace of the procedure rather than the pace of a network, sitting quietly inside the moment instead of interrupting it. The clinician should never be waiting on the software; the software waits on them.
The pod — EndoPod, PathoPod, TomoPod — is the information system and archive: an EIS, LIS or RIS that organises the studies and their patient metadata, encapsulated as DICOM and aligned to HL7 with a HIPAA-aware posture. Where the pod runs is the customer's call. Keep it fully on-prem, so the archive and information system stay inside the facility; or run it hybrid, an on-prem pod paired with a cloud pod, with edge and cloud kept in sync asynchronously so studies are reachable wherever they are needed.
The same architecture flexes from a single small clinic to a medium-to-large hospital, and it is AI-assisted-ready: pluggable, extensible models plug in gradually for specific tasks and can be customised per customer. You turn on what helps and leave the rest. Whatever the deployment, the read ends the same way — a clinician reviews the AI-assisted draft, edits it and signs it. The edge and the pod do the carrying; the clinician keeps the decision.
内視鏡・病理・放射線をまたぐエッジとポッド。
エッジツールはケアの現場でリアルタイムに支援し、ポッドは情報システムであり保管庫です — オンプレミスまたはハイブリッドで、組み込み可能なAIと、確認・署名する臨床医とともに。
- EndoEdge + EndoPod — real-time endoscopy AI and a full EIS.
- PathoEdge + PathoPod — whole-slide imaging and a pathology LIS.
- TomoEdge + TomoPod — an AI-native RIS and PACS.
How a read works
From capture at the edge to a signed report
From the moment a study is captured at the edge to the moment a report is signed, the path is the same across every modality — and a clinician is in charge at every point along it.
It begins with capture at the edge, on the equipment a department already uses. The edge tool assists in real time and records the study — and the patient metadata that belongs with it — into the local edge database. Nothing about the way a clinician works has to change for the assistance to be useful; the platform meets the existing instrument where it is rather than asking a team to adopt a new way of doing the job.
Then the study syncs to the pod — on-prem, and to a cloud pod too if you run hybrid — encapsulated as DICOM and aligned to HL7. The sync is asynchronous, so the edge never waits on the network, and the study becomes accessible in the web-based viewer wherever it is needed. Pluggable AI models, added gradually and customised per customer, screen systematically and draft a semi-automated report. They propose; they do not decide.
Review is where judgement lives. In the web viewer, the clinician brings context the study cannot hold — the patient, the history, the reason for it — and weighs the AI-assisted draft against it. Only after that human review is the report signed, as a structured result the next clinician can read and trust. Four steps, one principle running through all of them: the edge and the pod assist, and the clinician decides.
The family of products
One idea, expressed across every place imaging is read
CloudKites is not a single app. It is a family — three edge-and-pod imaging products on a shared foundation, Myro the agentic assistant that works across the clinical routine, and a set of free tools that put genuinely useful clinical AI in anyone's hands.
At the centre sit three imaging products, each an edge tool plus a pod — EndoEdge and EndoPod for endoscopy, PathoEdge and PathoPod for pathology, TomoEdge and TomoPod for radiology. They look different on the surface because the work they support is different, but underneath they share one AI-native foundation and one model: real-time assistance at the edge, an information system and archive in the pod, on-prem or hybrid by choice, pluggable AI, and a clinician who reviews and signs. Whichever modality a department starts with, it inherits the same way of working, and that does not get diluted as you add more.
Around those products is Myro, an agentic assistant that listens, sees and reasons across the clinical routine — from admin and nursing care to image and lab summaries and the work of a clinical intern — while always leaving the decision to a person. And beyond the clinical platform sits a deliberately free layer: MedPodGP for general practice and Emu for everyone, on-device tools we give away because useful clinical AI should not be gated behind a budget. Together they are one idea, expressed wherever imaging is read and care is delivered.
ニュースルーム
CloudKitesの最新情報
製品アップデート、お知らせ、臨床AIに関する視点 — チームから直接お届けします。
MedPodGP: faster, clearer documentation from the consultation
A refreshed ambient-voice flow that turns the consultation into draft notes, prescriptions, referrals, and handouts — all entirely on the GP's own computer.
Free, useful clinical AI — for GPs and for everyone
Why we ship genuinely useful tools for free and run them entirely on your own computer. A look at MedPodGP for ambient GP documentation and Emu for local, multi-modal AI.