How to help clients crawl, walk, then run with high-tech AI
By Thierry Hubert | 2 minute read | October 4, 2018
As companies tap into the power of artificial intelligence, there’s often an impulse to run from the very beginning. There’s a mistaken belief that AI can solve everything.
For this reason, we call our company Darwin Ecosystem because we look to the crawl, walk and run model—taking a tiered approach to help ensure transformation success.
We help clients explore how AI, IoT, cloud and blockchain can transform their work, especially through pattern detection.
We build solutions that augment the AI services provided by the IBM Watson and Cloud platforms, which we merge with our own technology and algorithms to help clients across a diverse range of industries.
Uncovering personality insights
For example, we combine Watson Personality Insights with our own technology to conduct natural language analysis that can shed light on personality traits and intent within given contexts. This has a range of applications.
In terms of human resources, our analysis can help companies place employees into the right job and raise awareness of any issues they may experience. This can help diminish the enormous cost of employee attrition.
We’ve performed experiments within schools using this technology to help identify which students are under duress—either from abuse or a mental health issue. These students can then be matched with a school social worker or therapist to get the help they need.
This technology has also been used to help with police recruitment. Personality Insight results were compared to FBI indices for optimal officers and those at risk. This score was taken into consideration with other hiring criteria.
Identifying miscoding in healthcare
Within the healthcare industry, the International Statistical Classification of Diseases and Related Health Problems (ICD) is a medical classification list that contains codes for diseases, symptoms, abnormal findings, etc. that pertain to a patient’s well-being.
Coding for the ICD is difficult. Miscoding is a common problem that can cause headaches and an enormous amount of losses in hospital revenue.
We combined natural language interpretation, neural network technology, machine learning and crowdsourcing to design an app that can help ensure a doctor’s note is properly coded—saving time and resources.
But we’re not stopping there. We’re looking to see how similar technologies can be applied across the healthcare industry to spur innovation related to administration, regulatory compliance and diagnosis, among others.
Giving voice to the voiceless
This ability to detect patterns informed our work helping a quadriplegic child with Tourette’s Syndrome communicate. Using pattern detection and AI, we were able to analyze the EEG signal from her brain and juxtapose it with her mom’s understanding of what the child wanted. Hence, we created the possibility for her to communicate without the help of her mother.
While this technology has its limitations, and took a long time to develop, it still creates possibilities for people with disabilities. Further, insight gained from this work on pattern detection is now informing another project we’re doing on financial predictive models.
No one industry corners innovation—and often innovation in one industry spurs transformation within another. Darwin Ecosystem is here to help make that happen.
Listen to Thierry Hubert explain how Darwin Ecosystem build solutions that augment the AI services provided by IBM: