Among the various leading-edge technologies impacting the business world — such as artificial intelligence, blockchain, and virtual/augmented realities — AI is likely the best known to most people.
IBM has a long-running ad campaign including regular TV spots touting its Watson platform, AI is prominently featured in entertainment even beyond the science fiction genre, and anyone who has interacted with Siri on an iPhone or Alexa via an Amazon home assistant device has been in direct communication with an AI construct.
How hot is AI? Per a MoneyTree report from PricewaterhouseCoopers and CB Insights, AI startups across the U.S. raised $1.9 billion in VC funding during the first quarter of 2018. That was a record amount for the segment and represented a 29 percent increase over the fourth quarter in 2017.
AI startups across the U.S. raised $1.9 billion in VC funding during the first quarter of 2018.
Artificial intelligence itself is a very broad term covering a wide range of tech, and by most definitions encompasses areas including expert systems, neural networks, natural language processing, fuzzy logic, and robotics. In practice, this can result in an application that quickly sorts and tags a huge amount of data while “learning” along the way and getting better at its task, such as image recognition. It can also be experienced when a digital assistant like Siri or Alexa “understands” a question you pose however you choose to frame the wording without having to conform to a preset command.
The North Texas area has no shortage of companies that are involved in AI in some capacity. Dallas-based RoboKind uses AI at the core of its robots that interact with and educate autistic children. RoboKind’s former Chief Information Officer Matt Stevenson left the company to become a deep learning and data science consultant and is currently working on developing a deep neural network for the Bina48 robot modeled on an actual person. In February, Texas-based Hypergiant announced its official launch as the self-described “office of machine intelligence for the Fortune 500.”
Here’s a look at a cross-section of North Texas companies implementing AI:
Founder and former CEO of Brainspace Dave Copps recently stepped down from the company less than a year after the Addison startup sold to cybersecurity company Cyxtera Technologies in a $2.8 billion deal. He took the time to speak with Dallas Innovates about the company and the current state of AI.
“Augmented intelligence is the science of radically altering the effectiveness of people through their interaction with machine learning and AI.”
Brainspace was founded in 2008 with its machine learning platform based around helping companies make sense of the unstructured data inherent in human language. Copps said the platform is now an industry leader in text analytics, e-discovery, digital investigations, and defense intelligence with customers among the Global 2000 such as the “big five” international consulting firms, global intelligence agencies, and legal service providers. The Cyxtera deal rolled up four security and analytics companies, including Brainspace, which were layered on an acquisition of 57 datacenters from CenturyLink.
What made AI attractive to Copps in the first place was what he described as the disruptive impact it will have in “virtually all markets” over the next 10 years. Brainspace has 14 patents on its technology, and Copps said the company excels in machine learning. He added Brainspace customers see a 90 percent increase in efficiency and productivity through the platform’s interactive visualization that creates a bridge between people and AI. An example of this in action would be a visualization transforming five million documents into a one-page navigable map he explained.
“The world is not yet at the point where AI can operate without us and we are definitely now in an area we call ‘augmented intelligence,’ which is where [Brainspace is] spending the vast majority of [its] time right now,”Copps said. “In short, augmented intelligence is the science of radically altering the effectiveness of people through their interaction with machine learning and AI. With the massive increases in the amount of data being produced in the world it is simply not possible for us as humans to consume, understand, or learn from the data without help anymore.”
“We are truly approaching a time where we could see the complete reinvention of business, the eradication of disease, and even potentially the expansion of human lifetimes by decades …”
Looking toward the future, Copps said machine learning and AI will eventually be part of every industry through what he called “automating automation,” likening the process to the industrial revolution which automated manual work and the computer revolution where knowledge work was automated.
Currently, AI is rapidly evolving with Copps pointing out that last year for the first time deep learning systems learned how to learn. This is a critical step because as AI continues to teach itself how to learn, we won’t be able to completely understand how AI is learning he said. This eventually means people and AI machines will have to undergo an evolution of trust.
“I think you have to take sides on AI. You are either dystopian or utopian. I tend to fall on the utopian side for sure,” Copps said. “I am informed by the dark side, but I think, with AI, we are truly approaching a time where we could see the complete reinvention of business, the eradication of disease, and even potentially the expansion of human lifetimes by decades with better, cheaper, and more widely available health care. I’m an optimist, bring it on!”
DocSynk is based in Irving and brings AI solutions across health care in three specific areas, including: population health management, revenue cycle management, and AI chatbots for patient care and patient engagement.
For population health management, DocSynk offers an AI-powered risk bundle that IDs patients at risk of chronic conditions and automatically recommends treatment and care management pathways. One example is identifying patients with pre-diabetes or metabolic syndrome in order to get those people on MDPP (Medicare Diabetes Prevention Plan) and dramatically reduce the expected cost of care for those patients.
It has an AI chatbot to manage care and engagement for cancer patients and provide a way for those people to handle operational and financial aspects of their care.
In the back office, DocSynk’s revenue cycle management AI platform automates key processes of contract and claims management to save both time and money. The company said it has demonstrated a 100 times performance improvement for large RCM (regulatory compliance mark) organizations for some processes in contract and claims management.
According to DocSynk, it has just scratched the surface of the potential for AI applications in health care with the next step being uncovering specific use cases that deliver clinical, financial, or operational value to patients and health care organizations. Some areas to watch, according to the company, include human-AI interaction through solutions like DocSynk’s chatbot, cancer immunotherapy breakthroughs enabled by AI, and the mainstreaming of AI in the business side of health care in the next five years.
Another health-care-related startup company is Lantern Pharma headquartered in the Uptown area of Dallas and bringing the iterative, big data crunching ability of AI tech to make pharma and biotech more efficient by eliminating the resource-wasting trial and error process.
“It won’t be long before we will see precision nano-robots fighting against tumors.”
Specifically, it is tailoring multiple promising drug programs targeted for specific cancer patients using what the company described as its adaptive “3R strategy” — rescue, reposition, and revitalize — for abandoned drugs through advanced genomics and AI. Lantern Pharma said the 3R process also is applicable for placing new drugs.
The strategy uses the RADR (Response Algorithm for Drug Positioning and Rescue) tech platform to combine AI with data closely matching real-world evidence to target patients most likely to benefit from the drugs. The process uses AI to make predictions on how a variety of anticancer drugs will respond in specific patients.
Lantern Pharma’s algorithms have proven to be successful with more than 80 percent accuracy in over 40 drug/tumor combinations, according to the company. It said RADR-AI was built on millions of data points and is continuously learning and improving. The tech accurately predicted drug responders in a blind test of patient clinical outcome for breast cancer treatment.
In drug development alone, Lantern Pharma sees AI coupled with advanced genomics making clinical trial design more precise with the result of improving the probability of achieving high response rates for drug candidates from the current 20 to 40 percent up to 80 percent and above with the added benefit of shorter timelines and lower costs. In the coming years, Lantern Pharma predicts AI will become part of multiple areas of personalized medicine including diagnostic testing, treatment guidance and therapeutic intervention, telemedicine, and even robots replacing physicians.
“It won’t be long before we will see precision nano-robots fighting against tumors,” said Lantern Pharma.
The Arlington-based workflow management company was founded by James Staud and Christopher McMurrough with an eye toward meeting an automation industry challenge by leveraging collaborative robots, machine vision, and machine learning.
AI is the core of Cloud 9 Perception’s business of automating visually-based systems for industrial applications through machine learning algorithms, 2D and 3D perceptual computing algorithms, and data analytics. Specifically, Cloud 9 implements a subset of AI, perceptual computing, to enable its tech and products.
Cloud 9 anticipates logistics will be a particular area of growth in the near term for its business.
Its approach is what the company calls a strategy of “heterogeneous application development,” meaning it tackles each business challenge with a collection of optimized algorithms for 2D imaging, point cloud processing, and deep learning along with emerging techniques from ongoing AI research.
Cloud 9 is operating in a growing business sector. The Association for Advancing Automation, an industry organization, released a report that found the North American machine vision market had a strong 2017 with 14 percent overall growth to begin the year, and Cloud 9 said the strong figures reflect that more companies are realizing machine vision systems are the future of automation. Cloud 9 anticipates logistics will be a particular area of growth in the near term for its business.
As AI increasingly becomes integrated into all areas of business, Cloud 9 made the point that early adopters will gain a competitive advantage simply by having more time to train their neural nets with more and better data.
VC-funded, Dallas-based startup Panamplify brings AI to marketing agency client reporting. It accomplishes this by utilizing symbolic systems AI to automate agency performance reporting which traditionally is an extremely manual process. A task that would typically take an agency hours, if not days, to assemble is turned into an “easy button from start to finish” that reduces agency workload to near zero, said Panamplify CEO Michael Pratt.
“Our reports self-assemble into intelligent, logical narratives,” Pratt said.
The underlying AI engine powering Panamplify’s product “understands” the role of each component in the report narrative it is creating and uses the available data to automatically produce an optimal report. Pratt said symbolic systems AI, also known as intelligent systems, is useful for this purpose because the domain of information around agency performance reports is known and manageable and governed under well understood best practices. He said intelligence systems also are applicable to business areas such as port operations and wealth management.
AI is no stranger to the marketing world. IBM Watson has an entire component based on marketing processes and Salesforce has its Einstein AI platform as part of its Marketing Cloud product. AI’s most common application in marketing is in data management such as collating and tagging corporate content libraries, churning through customer data to offer marketers predictive analytics on future campaigns, and even automatically creating and optimizing digital ads.
Pratt pointed out that while most of these AI marketing applications are focused on extracting insights and efficiently running ads, he believes Panamplify is the only company using AI for client report generation.