Orange Engine was originally built for a group of young doctors. It started quietly — one builder, one community, one engine designed to make the load a little lighter for people whose profession is to help others. And it did. And that quiet proof is a living testament that if it could do this for one community, it could do it for anyone it could ever reach. So here it is, for everyone.
The dual ingestion pipeline — PDF.js with automatic Tesseract OCR fallback — means it handles whatever a busy person uploads without making them think about file types. The rate limiter doesn't say "limit reached." It says "Let's give way for others." The heart beats because the engine is alive and because the people using it deserve something that feels that way.
The heart on the screen beats while it works. That part is not accidental. A real anatomical heart, rendered in the interface — beating slowly at rest, pulsing faster in orange when the engine is working. It beats because the engine is alive. And because the people using it deserve something that feels that way.
People who spend their days taking care of others — and then spend their nights studying so they can take better care tomorrow — are underserved by every major study tool in existence. Anki is powerful but requires configuration. Quizlet is built for students, not practitioners. Notion is a second job. AI chat interfaces hallucinate under sustained extraction pressure. No tool combined the intake of real clinical study material — typed notes, uploaded PDFs, scanned documents — with AI-powered card generation, structured complexity levels, and a human-centered design that acknowledged the weight of what its users carry.
For medical communities specifically — nurses, physicians, allied health professionals preparing for board exams, in-service training, or continuing medical education — no purpose-built, community-licensed, serverless AI study card engine existed. Orange Engine was built to fill exactly that gap, beginning with a group of young doctors in the Philippines.
Accepts raw study notes typed directly or extracted from uploaded PDFs and scanned images. Sends material through a secure serverless pipeline to a large language model engine, which generates structured study cards at a chosen complexity level — Standard Parse, Conceptual Deep Dive, Structural Breakdown, or Clinical High-Yield. Cards render in the browser and export as a formatted PDF. A real anatomical heart beats slowly at rest and pulses faster in orange when the engine is working. The rate limiter speaks in the language of the community it was built for: "Let's give way for others. It'll be your turn again soon."
Some people spend their days taking care of others — and then spend their nights studying so they can take better care tomorrow. This tool was built for them. You paste your notes or upload your document, tell it how many study points you need, and it turns everything into clean, readable cards you can actually retain. It's a helping hand for the people who are always helping someone else. The heart on the screen beats while it works. That part is not accidental.
Netlify serverless function intercepts POST — API key never exposed to client. LLM engine via /v1/messages, max_tokens 2048. Prompt enforces strict HTML output — no Markdown, no code fences, direct DOM injection via innerHTML. Dual ingestion: PDF.js v3.11.174 for digital text; Tesseract.js v5 OCR fires on pages with fewer than 30 extractable characters. Rate limiter: localStorage timestamp array, 50-call ceiling per rolling hour — communal by design. In-app browser jailbreak: navigator.userAgent detects Facebook, Instagram, Messenger — full viewport replaced on DOMContentLoaded. Heart: real anatomical photograph, mech-beat keyframe 1.8s cubic-bezier at rest; engine-active triggers mech-beat-fast at 0.5s with drop-shadow(0 0 20px rgba(255,94,0,0.9)) and brightness(1.2) on compile. Export: html2pdf() backgroundColor #020617, scale 2, print-color-adjust: exact. Talk to Barb: headless Google Form via hidden iframe tunnel, onload callback swaps form for success message, restores after 3 seconds.
The global medical education market was valued at $191.5 billion in 2024 and is projected to reach $678.6 billion by 2033 at a 14.9% CAGR (Business Research Insights, 2024). Within that market, the e-learning segment held 54.65% of the healthcare education solution market in 2025, while AI-driven learning tools are growing at some of the fastest rates in the sector (Mordor Intelligence, 2026). The AI in healthcare market specifically — which includes AI-powered education, clinical decision support, and research tools — is projected to grow from $36.67 billion in 2025 to over $505 billion by 2033 at a 38.9% CAGR (Grand View Research, 2025).
Despite this scale, the tools reaching frontline healthcare workers — the nurses, allied health professionals, and junior physicians studying on their own time, outside formal institutional frameworks — are generic. Anki, Quizlet, and AI chat interfaces were not built for clinical high-yield study requirements, continuous education mandates, or the compressed time windows of a healthcare professional who studies between shifts.
A 2024 Microsoft-IDC study found that 79% of healthcare organizations are currently utilizing AI technology, with an ROI realized within 14 months generating $3.20 for every $1 invested (Grand View Research, 2025). The investment case for AI in healthcare is established. What has not been built — until Orange Engine — is a community-licensed, serverless, zero-account, purpose-designed AI study card engine for the healthcare professional who learns alone, at night, so they can give more the next morning. Orange Engine was built for a group of young doctors first. It has been a genuine help to them — and it can be built to be that same help for anyone who needs it. The human brain can only hold so much at once. When the material is dense, the hours are short, and the stakes are high, cognitive load — the mental effort required just to process information — becomes the real obstacle between a person and what they need to learn. Orange Engine reduces that load deliberately. It takes raw, unstructured study material and returns it as clean, structured cards at a chosen complexity level — so the mind spends its energy on understanding, not on organizing. Less friction between the person and the knowledge. More room to actually retain it.
Most AI study tools are built for the product. This one was built for the person. The dual ingestion pipeline — PDF.js with automatic Tesseract OCR fallback — means it handles whatever a busy person uploads without making them think about file types. The rate limiter does not say "limit reached." It says "Let's give way for others." The heart beats because the engine is alive and because the people using it deserve something that feels that way.
No login required. No cloud involved. No account to create, no credentials to manage, no session stored anywhere. A healthcare professional can open it on any device and begin immediately — and when the tab closes, nothing remains. For people working with clinical study material, that is not a convenience feature. It is a structural guarantee that the content they process is never held by a system they do not control.
No account. No cloud storage. No analytics. No tracking. No cookies. Study content is never transmitted beyond the serverless function and never held on any external system. Rate limiting is client-side via localStorage. In-app browser protection blocks sandboxed social media environments where secure API routing cannot be guaranteed.
Orange Engine aligns with HIPAA Technical Safeguard requirements by architecture — not by certification, not by a Business Associate Agreement. No PHI reaches an unauthorized third party. No sensitive content sits at rest on external systems. The BAA is structurally irrelevant because no third party ever receives the content. The safeguard is the code itself.
Orange Engine was originally built for a group of young doctors. It has been a genuine help to them — and it can be built to be that same help for anyone who needs it. If your people carry the weight of learning so they can show up better for others — this engine can be made for them, in their language, with their community's heart beating at the center of it. Reach out via LinkedIn.