
"With new AI models, health tracking companies have realized that they can now provide insights using both structured and unstructured data. The new goal is to create interfaces and modalities that make it easier for users to create a habit of logging their meals or workouts, along with having an ever-present AI assistant that can guide people in areas like nutrition and exercise."
"The startup is using OpenAI's tech to power this conversational mode. With this release, Ria supports more than 50 languages, including 14 Indian languages. The company said that it can also support mixed language input like Hinglish or Spanglish. While the company is largely utilizing OpenAI's models for this release, it said that in the future it could use other models if needed."
"Through the new version of Ria, users can ask for their health overview for specified time frames like day, week, or month, or an overall summary. The app can pull data from different sources like fitness trackers, sleep trackers, or glucose monitors to give users insights about exercise, sleep, readiness, and glucose spikes, and give suggestions. Just like Google Gemini's Live Conversation mode, you can point the camera to ask about different food items and their nutritional value, then log them."
Healthify launched Ria, a conversational health assistant accessible via live chat, voice, and camera-based food input. Ria supports more than 50 languages, including 14 Indian languages, and accepts mixed-language inputs such as Hinglish or Spanglish. The assistant integrates data from fitness trackers, sleep trackers, and glucose monitors to produce insights on exercise, sleep, readiness, and glucose spikes and to offer suggestions. Ria can summarize health data over day, week, month, or overall timeframes, generate exercise plans, and log meals via camera or natural-language descriptions. The assistant currently uses OpenAI models and can operate with wearable devices like Ray-Ban Meta smart glasses.
Read at TechCrunch
Unable to calculate read time
Collection
[
|
...
]