Launching today

Spinachi AI
Your family's personal health manager on WhatsApp
27 followers
Your family's personal health manager on WhatsApp
27 followers
Your health, finally in one place. Forward your lab reports on WhatsApp. Spinachi sorts them for you and your family. See where you stand, how it's changing, and when your next test is due. No app. Never a diagnosis.



@chinmay3199 I couldn't agree more. I have built on Google Cloud Infrastructure and it was a breeze. GCP offers granular security controls that make configuration much easier.
Quick question: keeping health data in Mumbai region covers data residency, but how did you handle privacy beyond that? WhatsApp sits in the data path, so what stays on WhatsApp / Meta versus your own infrastructure, and do reports or data reach external APIs?
@richatsealedvault Thanks, Richard. That’s an important distinction, medical records are highly sensitive personal information, and protecting them is a core engineering priority.
We keeps a user’s reports and prescriptions together so they can retrieve records and track changes over time. Their dashboard presents those records, with a verification code sent to their WhatsApp making sure only the legit user gets the access.
We don’t use users’ health records to train models. However, that’s separate from processing, WhatsApp/Meta handles the messages and attachments exchanged with us. We are using Claude Haiku to process relevant information to provide the service. We store health facts like smoking, drinking etc to maintain continuity across conversations, those facts can form part of the context provided to our AI services.
Would love to take your insights on what you think more on this. Cheers
@chinmay3199 Thanks for the quick response, Chinmay. I appreciate your openness about the WhatsApp messaging boundary, the model usage, and the authentication trade-off. This kind of transparency builds trust.
Personally, I think balancing privacy with accessibility is genuinely hard, and I think you've built something unique. Once the data flow and retention are clearly explained to users, I expect anyone trying the product to feel confident. Congratulations and good luck with the launch.