Biofourmis founder emerges with new startup OutcomesAI; partners SingHealth
Tech firm aims to increase capacity of existing nurses and care teams, reducing their time spent on administrative clinical tasks by almost 50%
OUTCOMESAI, a healthtech startup, on Thursday (Oct 17) announced a partnership with SingHealth to integrate its artificial intelligence (AI) models with the health cluster.
The startup, founded by Biofourmis founder Kuldeep Singh Rajput, has built a multi-modal AI model, Glia, which can integrate information from images, audio and video to build a more accurate model. It is an ensemble of large language models and large multi-modal models geared for healthcare.
Rajput stepped down from biotech startup Biofourmis in August last year, a month after 120 employees were laid off. A year is a long time, said Rajput, who added that he had been running Biofourmis for about seven and a half years.
“The timing is right, (there were) a lot of learnings from Biofourmis on (the) healthcare ecosystem in general, on how to build and scale companies,” he told The Business Times.
OutcomesAI aims to increase the capacity of existing nurses and care teams, reducing their time spent on administrative clinical tasks by almost 50 per cent, according to Rajput. These tasks include chart reviews, triaging and check-ins with patients.
Glia can aid with real-time suggestions for healthcare professionals on what to assess and when to intervene, while also documenting consultations. This includes writing referrals, assisting with care coordination and education materials, said Rajput.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
The AI model has obtained about 25 nursing and medical certifications.
In addition, Glia has a panel of clinical teams to validate and test the models based on their speciality, from clinical reasoning to medication.
Only after these teams deem it to be safe, will the model be launched.
The current model is in its second version, with the first version taking at least three months to train.
The current model is focused on cardiology and oncology, with SingHealth to integrate Glia with the National Heart Centre and National Cancer Centre to enhance clinical workflows.
“We are leveraging its capabilities to not just improve workflows but to transform the very essence of patient care by improving access, timeliness and resource optimisation. In this way, our care teams can focus on what matters most: providing attentive and personalised care that focuses on each patient’s unique health concerns and needs,” said Professor Yeo Khung Keong, deputy group chief medical informatics officer (research), SingHealth, and CEO, National Heart Centre Singapore.
OutcomesAI wants to focus on multi-speciality areas, and started with cardiology and oncology due to market needs. Glia has models that are being trained for six other diseases, said Rajput.
Besides aiding in clinical workflows, OutcomesAI is working on a second product – an AI agent to do patient check-ins and care coordination.
This AI agent would be able to call and talk to a patient in multiple languages for symptom reviews, management on a weekly basis, taking over some nurse’s tasks and automating them.
Rajput added: “The rapid advancement of complex reasoning capabilities of AI and real-time voice technologies offer significant potential for enhancing patient engagement in non-diagnostic settings, such as intake, follow-ups, and even clinical trial follow-ups.”
Copyright SPH Media. All rights reserved.
