01 / Healthcare marketplace
MDsave
Clearer insurance information. Simpler data operations.
Patient-facing feature development · Internal data tooling
Case study draft
Executive summary
Among the many features I worked on at MDsave, this case highlights two separate projects: an end-to-end deductible checker for patients and a reusable Excel upload engine for internal data operations. The engine combined editable automatic column mapping, RabbitMQ background processing, and emailed XLSX files for failed-row recovery.
At a glance
- Deductible checker audience
- Patient-facing
- Upload engine audience
- Internal
- Deductible checker contribution
- End to end
- Upload engine: previous manual effort per hospital
- 20 hours
Behind the build
Technology stack
The development toolkit and connected services used at MDsave.
RabbitMQ supported background imports; PHPUnit and Codeception supported my unit and functional testing work.
Development & infrastructure
Building, testing, and delivering the product.
React
PHP
PostgreSQL
Redis
RabbitMQ
Jest
PHPUnit
Codeception
Ant Design
Storybook- WebSockets
CircleCI
AWS
Services & integrations
External services connected to the platform.
- Contentful
- Authorize.net
- Ribbon Health
- Prerender.io
- Salesforce
- Datadog
- FullStory
Introduction
MDsave is a healthcare marketplace connecting patients with providers and upfront procedure prices. I worked on many features within that business. This case focuses on two of those contributions, addressing distinct patient-facing and internal needs.
The deductible checker helps patients understand their insurance position. The upload engine helps employees manage large Excel files and add hospital data in batches. The engine is a separate internal tool; it is not part of the checker’s insurance lookup flow.
Feature summary
Deductible checker · Insurance details entry
The public form asks for insurance provider, first and last name, member ID, and date of birth.
Deductible checker · Deductible progress
The published experience shows individual and family deductible information, including what has been paid and what remains.
Deductible checker · A path into care discovery
The surrounding experience explains insurance terms and links to procedures and providers.
Excel upload engine · Automatic column mapping
Recognized spreadsheet columns are mapped automatically. Unrecognized columns can be assigned through dropdowns, and automatic selections can also be corrected.
Excel upload engine · Background processing
RabbitMQ supports processing large XLSX files in the background, with an in-progress status and an email when the operation finishes.
Excel upload engine · Failed-row recovery
Rows with missing required fields or other data issues are collected into a new XLSX file and emailed for correction and re-upload.
Excel upload engine · Reusable data workflow
The same reusable workflow could import different data types, beyond hospital additions.
The work
Project 01 · Patient-facing
Deductible checker
Start with the patient’s question
The feature answers a practical question: how much of my deductible is left? I worked on it end to end, bringing that question into MDsave’s patient-facing experience.
The public tool is available without signing up or logging in. It begins with a short insurance-information form rather than requiring patients to navigate a separate account area.
Make deductible progress understandable
MDsave describes the result as showing individual and family deductibles, the amount paid, and the amount remaining. These distinctions matter because a family plan can have both individual and family thresholds.
The case focuses on that published information flow. Insurance integrations and implementation details will be added when the technical walkthrough is documented.
Connect the result to a next step
The public page continues from deductible information into procedure and provider discovery. It also explains terms such as deductible, copay, coinsurance, and out-of-pocket maximum.
That context helps the checker function as part of the care-shopping journey: understand the insurance position, then explore available care and upfront prices.
Project 02 · Internal operations
Excel upload engine
Address the repetitive operational task
Adding a hospital meant manually entering all of its many data points. That detailed, field-by-field work took employees around 20 hours per hospital, and had to be repeated for each additional hospital.
The upload engine addressed the volume of manual data entry: it let employees bring those data points in through spreadsheets and process multiple hospital records together.
Recognize the columns, keep people in control
The upload engine automatically recognized spreadsheet columns and selected the corresponding fields. Employees did not have to configure every column from scratch.
When a column could not be recognized, a dropdown let the employee choose the correct field. Automatically selected mappings were editable too, so an incorrect match could be corrected before continuing.
Process large files without blocking the workflow
Large XLSX files were processed in the background using RabbitMQ. The user could see that the operation was in progress rather than waiting on the upload screen for the entire task.
An email notified the user when processing finished. This made the batch operation an asynchronous workflow with a clear status and completion notification. The 20-hour figure remains the historical manual effort per hospital, not a new processing-time benchmark.
Return the rows that need attention
During an add or upload operation, a row could have a missing required field or another data issue. The system collected the affected data into a new XLSX file and sent that file by email.
The user could correct just the problematic rows and upload them again. That made recovery a focused correction cycle instead of requiring employees to find the errors themselves across the original large spreadsheet.
Keep the engine reusable
Hospital data addition was one application. The engine was designed as a complete, reusable import feature that could add different types of data through the same workflow.
That gave the project a second outcome: beyond streamlining one workflow, it created tooling that could be applied to similar bulk data management needs.
Impact
Deductible checker: The delivered contribution is an end-to-end patient-facing checker. Its public experience provides a way to understand deductible progress without a login and continue into care discovery.
Excel upload engine: automatic column recognition reduced manual setup, RabbitMQ supported background processing of large files, and emailed failed-row spreadsheets made corrections and re-upload more focused. The previous hospital-addition workflow required around 20 employee hours per hospital.
Reuse was part of the result: the same engine could import different data types rather than serving only hospital additions.
Conclusion
The deductible checker and upload engine were separate projects within my MDsave work. One made insurance information easier for patients to understand; the other streamlined repetitive employee work through reusable data tooling.