Welcome Chris Diehl and David Gutelius, Co-Founders of The Data Guild, which create data driven solutions that tackle some of the world’s toughest challenges.
As Data Science proliferates, the ability to marry disparate sources of information and uncover relationships among entities continues to surface new solutions that was not possible even a decade ago.
Both Chris and David have extensive experience in designing and developing advanced analytics for companies like Jive, Proximial Labs, DARPA, and John Hopkins University to name a few. This episode was a pleasant learning experience for me as it allowed me to witness an altruistic side of Data Science that may be not have been readily apparent in this age of perpetual tech disruption.
In this episode we discussed:
- Business transformation: “Out with the old mindset”.
- The mission of the Data Guild and where it all began.
- Faer: The Data Guild’s data-driven solution for the health care industry.
- “Marketing is Spam. Real engagement makes people happy, fulfilled, heard and motivated.” David and Chris discuss their new product, Motion, and how this will tackle some of today’s challenges in marketing.
- What is the vision for Data Guild in the coming years?
You can listen to the podcast here or catch the episode on Libsyn.
Tackling wicked problems
This is what Data Guild sees as much as a challenge as it is a pleasure to conquer: Complex social issues – problems that we care about in the world.
David and Chris, coming from a variety of datatech experiences in government, health care and software, sought to create an organization that could focus on the tough challenges that crop up today and bring together disparate talents as they called “organized odd ducks” to pursue this.
Most problems are designed for diversity so the talent required people who did not only have specific talents in applied statistics, product design, but also diverse perspectives on the issues.
Data Guild started as a services organization but now is an incubator of product companies.
Organizations are not lock-in-step with the market evolution quickly unfolding
This is a clear symptom across organizations. The issue with archaic legacy systems and processes that inhibit more effective decision-making means organizations are not ready to embrace a data-driven culture.
Taking baby steps to introduce data science and measurement capabilities will help validate the need for new systems, however nurturing it throughout the organization will take time. As per Chris,
We can bring the capability but if companies are unwilling to leverage it completely and act upon it in a timely fashion, such that we make our way through the learning loop faster then its all for NOT.
David agrees that it is not a question of systems – but of people and culture that ultimately impede progress.
The more advanced usage of data means the decision making loop also needs to change. That impacts every facet of the organization. Nothing will change if the organization doesn’t change.
Organizations of the future will be built on the notion of organic decision making with data in the loop as a natural course of that growth – NOT something that’s bolted on as an afterthought.
Enter FAER: Patient triage for frontline community health care care staff
One of the applications Data Guild launched merged automation with direct human connection to help support a system already constrained with lack of resources. This, within an environment where chronic disease is unrelenting.
Data Guild developed a creative solution to transform a system that is unsustainable
Chronic disease accounts for some 80% of health care costs in the US, and growing. The current medical system was designed in the 19th century to address acute care, hence is dramatically ill-suited to the challenges faced today. The result is an overburdened, under-performing medical system built on a broken economic model, propped up for the moment by inertia.
In Chicago, where Faer was piloted, there were 24 case workers among a patient population of 6,000. That is a ratio of 250 patients per case worker.
Faer needed to be a solution that dealt with the resource concerns of the community of health care care workers and their current environmental context. At the same time Faer needed to address the patient’s need: for support and access to information.
These class of problems came down to this: Ensuring value based-care that is properly coordinated with the goal of improving outcomes for the patients and the case workers.
The Solution: Faer became a part of the care team. It filled a gap to allow caregivers to allow patients to manage their conditions. Machine learning was a way to learn the needs of that patient in real time. The team incorporated data including full medical history, environmental data as well as claims history to ensure ease of access. Using SMS as a communication mode between patient and caregiver, the system learns the “context-specific recommendations” within this dialogue and develops a greater understanding of the need and the optimized response. It will become proficient in the kind of supports and interventions needed to produce positive outcomes for patients (tailored to each needs).
This is not easy… clearly.
Faer is always on and always available to patients. It can automate many tasks like appointments, follow-ups, and direct access to care team. But it also relies on a cloud-based engine combining “big data with proprietary machine learning, behavioural science, and clinically validated intervention design”.
Ultimately, this notion of self-care can dramatically improve individual case load while giving patients some level of control and peace of mind.
Are practitioners ready for this? Adoption again is key. It will be interesting to see how a system like Faer can be applied to other health care challenges as the Boomer population ages.
Motion: the marketing solution to dealing with SPAM
For two scientists without a marketing background it was refreshing to see their point of view on a topic that continues to be the bane of this industry.
I read this on the presentation deck for Motion, a new venture for The Data Guild:
Real engagement makes people happy, fulfilled, heard and motivated.
The insight came about when The Data Guild was helping a client rethink the patient billing process. Through a marketing automation platform the one way push communication did not take into consideration the patient context, which ultimately would impact the business longer term.
As per Chris,
People were annoyed by it. It is not the way it could or should be. Establishing some level of trust is good marketing to nurture the long term relationships.
As David and Chris continued to discover: a single click did not necessarily denote acceptance.
Marketing automation has become so readily accepted that it misses the point of building high trust
In the process, this causes a rift between the organization and its customers.
Building systems from the standpoint of empathy by watching customer behavior, how and what they respond to is the holy grail of response marketing. But it needs to be extended throughout the lifecycle of the customer relationship.
It means changing perspective from “who will buy” to “why they buy” and “why they continue to buy”.
It becomes less about their demographics and who they are as customers, but rather, who they are as people. As David notes,
Ad tech was built on a system of spying, logging people’s behaviors and assuming affinities. There is a better way that’s more human. Start from a place of respect, a place of empathy.
While today, solution bots are everywhere. Chris emphasized there needs to be the right balance where humans need to be part of that mix – to deliver value.
Where do humans come into the mix? How do we present those things to humans and make that experience more compelling so it provides value. The more you hand off to machines, it effectively dehumanizes the conversation.
It is so green field but we are both excited. It is the realm of the possible: the right composition of human and machine interactions… Machines learn the pathways of communication and meaning to help people in organizations leverage that information in positive ways.
Machines are unable to mimic human decision
This is consistent with all of the data scientists we interviewed in this series.
Both were adamant: “Why are we trying to endow machines with human capabilities when we need to empower humans with machine capability (through enhancing natural intelligence)?”.
Humans have cognitive limitations. Only machines have the cognitive load to aggregate and find patterns within the information.We need to work with tools that run on machines that can help humans respond to the tough problems. Only humans have the emotional capability to do this today.
Designing for the future
David’s long term plan is to have the Data Guild become the engine for solving many of these problems at scale and having a true impact on the world. Today they have the prototyping capability with lots of problems currently being tackled.
For Chris, the foundation of the Guild is the community. Within 5 years time he wants to continue to nurture a tribe of people trying to make a difference in climate change, health care, energy etc.
There are many questions about the application of the technologies and their implications. As per Chris,
We need more people who are mindful of those implications and be willing to say no to certain applications of the technology… The arms-race mentality is dangerous: If we don’t build it, someone else will.
We need to be more empathic to the broader concerns and how technology may help. More importantly, the best way we may help is NOT to intervene. Building a tribe that believes that can make a significant difference in the world.
About David Gutelius: Dave previously drove the data science and technology innovation at Jive Software as a Chief social scientist. He also founded Proximal Labs (a machine learning company) which was acquired by Jive. David also cofounded the Social Computing Group and while there, he worked with client, DARPA (the emerging technology arm of the US Dept of Defence) on one of the largest machine learning projects in US Govt history called CALO.
About Chris Diehl: Chris has extensive experience defining and developing analytics for different sense making and prediction tasks. Chris was also the principle scientist at Jive Software, designing and developing advanced analytics for social enterprise and online community health assessment. Prior to that he spent 10 years at John Hoppkins’ University physics lab developing machine learning approaches for Department of Defence and intelligence community.
About Humanity in Data: Data isn’t only a story of numbers … or information … or insight. We forget, it is also about the people who interpret that data. Looking for Humanity in Data seeks to explain the ways in which behavioral data represents the mind and emotions of those who generate it. We do this by engaging with data science specialists and practitioners in this space. Our purpose is to translate from the scientific to the every day impact of technology on the individual. We’ll discuss topics about information collection, privacy and governance, increased contextualization and what is at stake for everyone.