Fictional panels
Two fixed samples are generated in the browser for demonstration.
Project #5 · OCR-Powered Health Data Processing & Educational Insights
Life Gauge
Extract fictional lab values with OCR, review and validate them, and explore an educational result in the browser.
Synthetic demonstration only. Not a medical product, diagnosis, or clinically validated prediction.
01 · Project identity
A visible health-data workflow, built with fictional inputs.
Life Gauge demonstrates a web-based workflow that extracts nine values and units from a fictional panel, presents them for review, validates the inputs, and calculates an educational result. An optional Gemini insight layer is integrated with a non-AI fallback; successful live AI generation has not been verified.
02 · Hero proof
From fictional data to clarity.
The V1.1 demonstration reads a fictional panel with browser OCR, displays confidence and editable review fields, and calculates an educational PhenoAge output. In the captured sample B run, simulation age 28 returned 17.8 years. This is a software output, not a health assessment.
03 · Problem
Make an opaque process easier to review.
When extraction, validation, and calculation happen out of sight, it is difficult to understand how a result was produced. A safe demonstration needs to make those stages visible without relying on real health information.
04 · Solution
Show each step with a constrained synthetic demo.
The browser generates a fictional panel and reads it with OCR. Visitors review editable detected values and units against the separate fictional reference, select a simulation age, and submit the reviewed values for validation and deterministic calculation. An optional Gemini educational insight is integrated; if it is unavailable, a non-AI fallback is shown. Only bundled fictional panels are supported.
05 · Workflow
Choose, read, check, and calculate.
01 · SelectChoose fictional sample A or B
02 · GenerateCreate its panel in browser memory
03 · ReadRun bundled browser OCR
04 · ReviewEdit values and units; choose a simulation age
05 · CalculateValidate inputs and display the result
06 · Key features
Clear boundaries around each stage.
OCR and validation
Browser OCR displays confidence for detected values and units. Editable review cards keep OCR evidence separate from the fictional reference; the server validates reviewed inputs before calculation.
Educational output
A configurable simulation age feeds the deterministic calculation. Optional Gemini educational insight is integrated with a non-AI fallback; successful live Gemini output has not been verified.
07 · Visual proof
The V1.1 workflow, on desktop and mobile.
These real captures show fictional sample B, nine OCR review cards, and a 17.8-year result at simulation age 28. The result is not a medical assessment, and the images do not claim a successful Gemini response.
Scope: Fictional data only. No real reports, diagnosis, medical advice, clinical validation, or arbitrary report compatibility is claimed.
08 · Business value
A workflow stakeholders can inspect.
Visible stages
See how extraction, checks, calculation, and explanatory copy fit together.
Reviewable boundaries
Bundled fictional panels and explicit limitations keep the demonstration focused.
Reusable potential
The workflow illustrates components for a Delivery Kit candidate, subject to further packaging and validation.
09 · My role
Design and implement the demonstration.
Designed and implemented the synthetic browser experience, OCR review and validation flow, calculation and educational output presentation, and release evidence.
10 · Tech stack
Practical components with clear boundaries.
The reusable candidate spans OCR processing, input validation, biomarker normalization, calculation integration, configurable workflow and adapters, and a synthetic demonstration environment. The optional Gemini integration has a non-AI fallback; live AI-generated success remains unverified. It is not a packaged or clinically validated medical SaaS.
HTML · CSS · JavaScriptTesseract.jsDeterministic validationCloudflare WorkersOptional Gemini integration
11 · Proof links
Explore the synthetic demonstration.
The public web demo runs without an account or saved submission. Use bundled fictional samples only. Gemini-generated output is not yet verified in the live environment; a non-AI fallback is available when the service is unavailable.
12 · Next step
Need a clearer document or data workflow?
Let’s map your inputs, validation rules, exceptions, and human review points.