Sully.ai
@sully_the_ai
Followers
589
Following
17
Media
11
Statuses
52
Autonomous OS for healthcare
San Francisco Bay Area
Joined December 2017
From 0 to 400+ healthcare organizations & counting🚀 https://t.co/0bmP94nhPj’s AI Medical Agents are powering clinical operations—streamlining workflows and freeing clinicians to focus on patients ⚡️ #HealthTech #DigitalHealth #AIinHealthcare #HealthcareInnovation #HealthIT
4
1
11
Excited to announce our partnership with evolvedMD, one of America's fastest-growing behavioral health companies: deploying an autonomous operating system – enabling clinical services and outcomes to support more patients annually across 200+ primary care sites.
3
0
6
Go team @sully_the_ai! đź’Ş
2
3
13
🎯
We’re hiring engineers + eng leaders at https://t.co/ep0hinFnQw! If you haven’t built and shipped at a startup before — this isn’t for you! Ex-founders, MDs & researchers from MIT, Stanford & NASA; building real AI agents used in hospitals today to help doctors save more
1
0
1
You can't box me!
0
0
1
autonomous OS ↓ autonomous scheduling ↓ autonomous lab orders ↓ autonomous documentation ↓ autonomous data-entry ↓ autonomous patient follow-ups
0
2
3
I'm gonna try my best...
at Sully, we're building the Defacto of healthcare AI. Just shipped receptionist and voice agents that are now in customers hands, making @sully_the_ai the first company on the planet to have receptionist agents connected with scribe and other agents.
0
0
2
Yup, in healthcare, https://t.co/0bmP94nPER always wins!
sully.ai
Sully.ai saves doctors time with an AI team: Nurse, Receptionist, Scribe, Med Asst, Coder & Pharmacy Tech—seamless handling check-in to prescriptions.
GPT-5 on MedXpert was 53%, still our consensus mechanism blow them out of the water by like 8% higher (61%). ArXiv:
0
0
2
MedXpert (Medical) Results for @OpenAI GPT-OSS 120B OpenSource Model Final Accuracy: 32.20% Accuracy by Medical Task: Basic Science: 37/130 = 28.46% Diagnosis: 72/213 = 33.80% Treatment: 52/157 = 33.12% Accuracy by Question Type: Reasoning: 133/390 = 34.10% Understanding:
0
3
8
Model Accuracy: Hallucinated vs Ground Truth in Medical Runs (80–100% Scale) Highest: Claude-opus ~95% Lowest Gemini 2.5 Flash ~83%
0
1
1
Clinicians at Nile Women's Health now reclaim 2.4 hrs/day from admin tasks using https://t.co/0bmP94nhPj’s AI Medical Agents. See how https://t.co/0bmP94nhPj transformed Nile’s workflow👇 https://t.co/fffNmy63ym
#HealthTech #DigitalHealth #AIinHealthcare #HealthcareInnovation
sully.ai
Sully.ai saves doctors time with an AI team: Nurse, Receptionist, Scribe, Med Asst, Coder & Pharmacy Tech—seamless handling check-in to prescriptions.
6
2
6