Brahva

Inspiration

Healthcare pricing in the U.S. is technically “transparent” due to CMS regulations, but in practice it is nearly impossible for patients to understand or compare costs. Hospitals publish massive, inconsistent machine-readable files that are not usable for everyday decision-making. We were inspired by this gap between data availability and data usability, and wanted to build a system that turns mandated transparency into real-world clarity for patients.

What it does

Our platform helps users estimate and compare the cost of medical procedures across hospitals. A user enters a procedure (e.g., MRI, CT scan), selects their insurance provider and insurance type, and our system returns estimated cost ranges across multiple hospitals.

For this demo, we use a variety of hospitals across the Philadelphia and New York City area. The system normalizes CMS hospital pricing data, maps medical procedures using standardized codes, and generates side-by-side cost comparisons so users can quickly identify the most affordable option.

How we built it

We started by intensely brainstorming startup ideas and narrowed them down to two strong options. We chose healthcare price transparency because it was the bigger problem with real impact. From there, we challenged the idea’s feasibility, risks, and upside together, using AI tools to help pressure-test assumptions and move faster.

We then split into focused roles. Two team members led coding, frontend, backend, and GitHub collaboration. Two focused on operations, UX, product direction, and understanding how healthcare pricing actually works so we could solve the right problem.

Claude, ChatGPT, and Claude Code became major force multipliers throughout the process. They helped us code faster, debug issues, refine UX, and execute quickly. But the ideas, decisions, leadership, and passion, came from our team. We used AI to accelerate execution, not replace thinking.

We built in parallel, constantly checking in, iterating, and improving together. As we neared the deadline, we shifted into polish mode: refining the product, completing deliverables, and mapping ways to grow this beyond the competition into a real nationwide platform.

See the github repo README for the full tech stack.

Challenges we ran into

One of the biggest challenges was dealing with the inconsistency of CMS hospital data. Each hospital publishes its files differently, with varying structures, naming conventions, and levels of detail.

Another challenge was designing a pricing model that is both useful and honest. Insurance billing is highly complex, so we had to balance usability with realistic assumptions without overclaiming precision. One compromise we made was not using a patient's deductible information to form a price because this would lead the outcome from being a price estimate to a personalized price, which could potentially be false or misleading due to the variety of factors that affect what the finalized billing price could be.

Finally, mapping procedures across hospitals required careful normalization using CPT/DRG codes to ensure accurate comparisons.

Accomplishments that we're proud of

We successfully transformed raw, complex CMS hospital pricing data into a structured and usable comparison tool. We built a system that allows users to move from uncertainty to clarity in seconds when comparing healthcare costs.

We’re especially proud of creating a workflow that feels simple to the user while handling highly complex data processing behind the scenes. We also designed a system that is scalable beyond our demo dataset and can be extended to hospitals nationwide.

What we learned

We learned how difficult healthcare data standardization is in practice, even when federal regulations require disclosure. We also gained experience working with large, inconsistent real-world datasets and building normalization pipelines to make them usable.

On the product side, we learned how important it is to translate complex data into simple, actionable decisions rather than exposing raw information.

What's next for Brahva

Next, we plan to expand beyond our initial twelve hospitals and scale nationwide using CMS pricing data while improving personalization through better insurance modeling, broader procedure coverage, and stronger cost estimates. We also aim to build an automated backend pipeline that continuously scans for new hospital pricing files, updates data, and expands coverage over time. Future versions could include verified user feedback so patients can report actual billed costs to improve estimate accuracy and trust. Long term, we want to combine pricing with quality-of-care metrics to become a complete healthcare decision platform

*Currently in demo mode for select Westchester and Philadelphia hospitals.