Second Opinion

Inspiration

Americans over 60 lost $7.7 billion to scams in 2025, and people with disabilities are defrauded even more often, so we built a user-friendly scam checker intended to detect and prevent scams.

What it does

Provides information and guidance, regarding the input of text or image, in a simple way for the elderly and disabled.

How we built it

A Python/FastAPI backend combines a scam-pattern rules engine with the Anthropic API, connected to an accessibility-first HTML/CSS/JS frontend

Challenges we ran into

Our team had trouble connecting the frontend and backend in parallel, resolving Git merge conflicts, and making sure the AI was able to receive the frontend returned inputs

Accomplishments that we're proud of

A fully deployed website with a successfully integrated AI feature that allows for in-depth analysis based on user input

What we learned

How to design APIs for a team, use AI safely, collaborate with Git, and let our users' needs drive every design choice.

What's next for Second Opinion

Making it universally accessible: different languages and better AI tools.

Built With

Submitted to

Rowdy Hacks XII

Created by

I proposed the idea for Second Opinion, a plain-language scam checker built for the elderly and disabled. I also led its backend. I built the core scam-detection engine, which combines pattern-based safety rules with the Anthropic API to explain red flags in plain language, score each message's risk, name the scam type, and clearly identify the spectrum of scam likelihood. I also set up and managed our GitHub repository, integrated the team's frontend and backend into one working app, wrote the tests, and deployed the full site on Render.

AustinOkuno
I worked on the sprites and the animation for the mascot we created, Scampy! Along with our adorable racoon, I also strived towards fine tuning our existing website by creating a stress free UI that targeted towards making a more elderly accessible website

gtms
I worked front-end and effectively pitched the team's idea to the judges.

Mars Marselin
I worked on backend API using FastAPI and was responsible for its reliability, meaning what the frontend can send, validating text and image inputs, handling failures and writing test cases for any possible failures.

Naomi Teklemichael