CredAI
CredAI

Role
Proposal and scope
Team
Dev team
Platform
Web App
Timeline
Ongoing
project overview
CredAI is an AI-powered web platform that automates healthcare credentialing: it reads provider documents, fills in and submits insurer applications, and tracks them live. I turned the client’s one-line idea into an approved scope and proposal.
CredAI is an AI-powered web platform that automates healthcare credentialing: it reads provider documents, fills in and submits insurer applications, and tracks them live. I turned the client’s one-line idea into an approved scope and proposal.
The ask
A founder wanted an AI-run service that handles insurance credentialing for healthcare providers, with as few people involved as possible. Credentialing is the paperwork providers must complete before insurers will pay them, and it usually takes around 90 days. He had no feature list and no similar product to compare against.
What we planned
I started with six questions on users, what the AI should do, the first version, and platforms. From his answers I mapped the full journey: the provider signs up, uploads documents, the AI reads them, matches insurers, fills in applications, submits them, tracks status and follows up automatically.

Key decisions
Honest on speed. He wanted approval in 7 days instead of 90. Insurers control approval time, so I told him we can't promise that. We focused on what we can control: faster submissions, fewer rejections, and prioritizing quicker insurers.
Getting access. He had no access to the systems we needed. I explained how developer accounts work, which are instant and which need an application, and listed what he needed to provide.
Strengthening his rules. He sent his own rules for the AI, such as pausing on expired licenses. I kept them and added a human override, confidence scores for doubtful decisions, and per-insurer rules.
Making the demo a priority. He wanted a demo that sells the product, so a demo dashboard and marketing site were in scope from day one.


What was included
Provider onboarding, document upload with AI reading, insurer requirements, automatic form filling and submission, live status tracking, automated follow-ups, admin panel, marketing site and demo dashboard.

The ask
A founder wanted an AI-run service that handles insurance credentialing for healthcare providers, with as few people involved as possible. Credentialing is the paperwork providers must complete before insurers will pay them, and it usually takes around 90 days. He had no feature list and no similar product to compare against.
What we planned
I started with six questions on users, what the AI should do, the first version, and platforms. From his answers I mapped the full journey: the provider signs up, uploads documents, the AI reads them, matches insurers, fills in applications, submits them, tracks status and follows up automatically.

Key decisions
Honest on speed. He wanted approval in 7 days instead of 90. Insurers control approval time, so I told him we can't promise that. We focused on what we can control: faster submissions, fewer rejections, and prioritizing quicker insurers.
Getting access. He had no access to the systems we needed. I explained how developer accounts work, which are instant and which need an application, and listed what he needed to provide.
Strengthening his rules. He sent his own rules for the AI, such as pausing on expired licenses. I kept them and added a human override, confidence scores for doubtful decisions, and per-insurer rules.
Making the demo a priority. He wanted a demo that sells the product, so a demo dashboard and marketing site were in scope from day one.


What was included
Provider onboarding, document upload with AI reading, insurer requirements, automatic form filling and submission, live status tracking, automated follow-ups, admin panel, marketing site and demo dashboard.
