50% of time of ABA therapy is writing reports.
We combine personalized therapy with cutting-edge AI to track and prebuild progress reports, offering 50% more time for your child treatment.
.png)
Step 1
Automatic data collection
We record sessions, so our clinical team is always present for you child without interrupting to data collection and writing report notes. All data is securely stored, anonymised and deleted after analysis.
Step 2
Analyse functional behaviours
and create progress reports
Our behavior care engine performs functional behaviour analysis for target behaviours and skills and creates automatic progress reports for review and approval by our clinicians.
Step 3
Constantly improving treatment plans based on progress report data
Achieving notable progress is observed in 90% of children when recommended hours are fully implemented with active caregiver involvement.
Our behavior care engine updates behavior intervention and skill acquisition plans after each session based on achieved progress.
Our behavior care engine updates behavior intervention and skill acquisition plans after each session based on achieved progress.
FAQ
Is all the technology used HIPAA-compliant?
We anonymise and un-identify data before processing by LLMs and store it in accordance with HIPAA regulations.
Which specific model/version of each LLM you are using?
We use LLAMA 3, GPT4v, Gemini 1.5, Claude Opus 3, Mixtral 8x7b. Currently, we will be testing out the best models for each use cases (generate reports, generate insights, etc). So we use mostly all at the moment.
What is your policy around using inputs/outputs to train models?
We use anonymised inputs and outputs to finetune models. We won't store original videos we record. We basically take the video, anonymise it, process it and delete the original video.
Can we request that the data is deleted at the completion of our services?
Yes, we will delete data on your request.
We anonymise and un-identify data before processing by LLMs and store it in accordance with HIPAA regulations.
Which specific model/version of each LLM you are using?
We use LLAMA 3, GPT4v, Gemini 1.5, Claude Opus 3, Mixtral 8x7b. Currently, we will be testing out the best models for each use cases (generate reports, generate insights, etc). So we use mostly all at the moment.
What is your policy around using inputs/outputs to train models?
We use anonymised inputs and outputs to finetune models. We won't store original videos we record. We basically take the video, anonymise it, process it and delete the original video.
Can we request that the data is deleted at the completion of our services?
Yes, we will delete data on your request.