Hubble
Retrieve medical records other APIs can't
339 followers
Retrieve medical records other APIs can't
339 followers
Medical records still live behind fax lines, phone trees, and portal logins nobody remembers. A patient verifies their identity once, and Hubble assembles their records from across their providers and returns them through one API, for your AI agent to use.
Command AI
awesome product - congrats team!
Hubble
@vinayyala appreciate the support!
Refocus
The fallback chain is the interesting part: API first, then browser agents through provider portals, then voice agents calling records departments. I build voice AI in the aging-in-place space, daily check-in calls with older adults, so I have some scar tissue around that last step. Records departments tend to be IVR trees, long holds, and a human who ultimately wants a fax number.
Two questions. Does the voice agent disclose it's an AI to the clerk up front, and did that move your success rate either way? And when two sources return the same encounter with conflicting data, does the API hand back both versions or reconcile them for you?
Hubble
@igorgurovich great q's!
1. we always maintain upfront transparency that it is an automated caller on behalf of the patient and additionally disclose possible recordings and transcripts in two-party consent states. it can certainly be awkward at first, but we believe (a) people are increasingly accepting of multimodal communication with AI and (b) a learning automated caller can be more efficient to respond to at scale. we're still gathering insights, but happy to share notes once we find out more!
2. in theory that specific example shouldn't happen, but our current stance is to avoid lossiness, so we haven't formed a strong stance yet on reconciliation
EverTutor AI
Congrats on the launch
The problem you are solving is genuinely interesting especially with patient records being scattered across providers
Curious to see how Hubble handles the messy real world edge cases here and excited to see where this goes
Hubble
@suryansh_tiwari2 100%! There are a ton of challenging edge cases that make the problem of record retrieval an annoying one to build and maintain in house. It's the reason why we're investing in being that trusted bridge so that companies don't need to build this themselves.
EverTutor AI
@prabhadublish Exactly this 👌
The edge cases are probably where things get really interesting
Curious to see what you guys uncover as Hubble scales
Hubble
@suryansh_tiwari2 Appreciate the note! Will definitely share more as we progress.
EverTutor AI
@prabhadublish super excited! Wishing you a great launch day
Macaly
fax lines and portal logins are the real bottleneck here. nice work 🎉
Hubble
@petrkovacik so true!!