What a smart approach with the subpath import for v4, easy gradual adoption:

x.com/colinhacks/...

The Imperative of Smarter Clinical Information Systems How AI Agents and openEHR Can Lift Healthcare Out of Its Purgatory

👉 Full deep‑dive here: dub.sh/8VvOUfN

@openehr.bsky.social #healthIT #SwissHealth

Good news: AI Agents × openEHR kill that workflow tax.
Notes dictate themselves, PDFs self‑file, forms pre‑fill—clinicians go back to medicine.

Link below 👇🏽

Swiss healthcare now burns >12 % of GDP, yet half our junior docs say they “can’t go on.”
The silent culprit? Admin overload.

🔎 114 min/day: doctors stuck on keyboard, not bedside
🔥 74 % nurse burnout risk when paperwork >2 h/shift
📠 Yes, we still re‑type data from faxes in 2025

Byte-sized Brainwaves | Gordon Mickel | Substack Insights on Software Development & AI. Exploring AI Agents, Code Assistants, HealthTech innovation, OpenEHR integration, and the future of technology. Join us to ignite curiosity. Click to read Byte-sized Brainwaves, by Gordon Mickel, a Substack publication. Launched 2 years ago.

👉🏽 bytesizedbrainwaves.substack.com/

Illustration of two overwhelmed clinicians buried under flying paperwork; overlaid callouts show ‘114 min admin per doctor’, ‘74 % burnout risk’, and ‘16–19 % avoidable waste’, highlighting the crushing weight of administrative burden in hospitals

Tomorrow at 09:00 CET (on my blog): How modern hospitals turned clinicians into digital scribes—and why AI + openEHR will end that era. 👀

2nd openEHR Switzerland Symposium During this symposium, we delve into the strategic value of openEHR in the future development of eHealth solutions, exploring its significance, real-world applications, and the imperative need for industry standards and secondary use of health data.

My talk: "From Chaos to Clarity: Transforming Unstructured Data into openEHR with AI Agents." I've been elbows-deep in prototypes and can't wait to swap ideas. Curious? Details 👉🏽 aws-experience.com/emea/de-cen...

Hey #HealthIT friends—any @openehr.bsky.social folks around? I’ll be speaking at the 2nd OpenEHR Switzerland Symposium on 5 May in Zürich 🇨🇭. If you’re there, let’s grab a coffee and geek out on clean data! #OpenEHR

We offer flexible work arrangements, a collaborative environment, and the chance to tackle real-world healthcare challenges.

DM me for details or apply here: www.cistec.com/ai-engineer/

If you're passionate about:
✓ Applied AI in healthcare
✓ Building production-ready ML systems
✓ Full-stack development (TypeScript, Python, Node.js, React)
✓ Machine Learning, LLMs, and NLP
✓ MLOps/LLMOps and Vector Databases

• Next-gen AI capabilities for Switzerland's leading clinical system
• Smart tools for clinical documentation & data insights
• Intelligent interfaces that understand medical context
• Production-ready AI that delivers real value to medical professionals

At CISTEC, you'll be part of a passionate, interdisciplinary team at the forefront of AI innovation in healthcare. We're not just building tools, we're reimagining what a clinical information system can be.

What you'll work on:

We're building cutting-edge AI solutions for Switzerland's leading clinical information system, directly impacting patient care by reducing administrative burdens on medical professionals.

🏥 Calling all AI Engineers in Switzerland! Want to help reimagining healthcare workflows? Join my team at CISTEC!

This is one component of the KISIM AI platform. Still really excited about the challenges at the intersection of healthcare and AI.
#HealthTech #AI #MedicalAI #DigitalHealth

It's not just prompt engineering - we've built a complete pipeline for reliable clinical documentation generation. The system integrates with existing workflows while maintaining full transparency and traceability.

Quality Assurance Pipeline:
• Real-time evaluation through LLM judges
• Continuous quality scoring
• Automated accuracy ratings
• Live validation against source data
• Performance monitoring across different report types

Our system generates reports in any target language while preserving clinical accuracy and documentation standards. Each report maintains complete source reference integrity regardless of target language.

Healthcare providers can define custom generation patterns through templates and gold standard examples, allowing them to match existing documentation styles and implement department-specific formats.

First, the core architecture - source verification:
Each generated passage is linked to structured clinical data through an interactive reference system. Hover tooltips expose the exact source data and transformation path.

🧵 Been working on solving trust and verification challenges in medical AI. Here's our approach at CISTEC to reliable report generation:

Sure, sent you the last one, have fun, Kagi is great!

This project is a natural extension of my previous work, CodeWhisper, but with a more focused scope. I'll attempt to document each step and share them upon completion.

Going beyond "screenshot to frontend clone" functionality, it's quite trivial to setup an app that generates prompts for automated feature development for any full-stack web app stack using PRDs, Flow docs, DB schema, screens and a good understanding of how the project is setup.

I love how @CopyCoderAI uses Cursor/Bolt as an implementation agent or code orchestrator. I began developing my own version for our product development at @cistecag, using our tech stack (a large monorepo with NestJS, GraphQL, and React). The results are impressive.

I have two 3-month Kagi Search subscriptions to give away for free to interested followers. Hit me up—first come, first served!