Artificial Intelligence | News | Insights | AiThority
[bsfp-cryptocurrency style=”widget-18″ align=”marquee” columns=”6″ coins=”selected” coins-count=”6″ coins-selected=”BTC,ETH,XRP,LTC,EOS,ADA,XLM,NEO,LTC,EOS,XEM,DASH,USDT,BNB,QTUM,XVG,ONT,ZEC,STEEM” currency=”USD” title=”Cryptocurrency Widget” show_title=”0″ icon=”” scheme=”light” bs-show-desktop=”1″ bs-show-tablet=”1″ bs-show-phone=”1″ custom-css-class=”” custom-id=”” css=”.vc_custom_1523079266073{margin-bottom: 0px !important;padding-top: 0px !important;padding-bottom: 0px !important;}”]

AiThority Interview Series with Eric Dirst, COO, ThingLogix

ThinkLogix

Know My Company

Tell us about your technological interaction with IoT and other intelligent technologies that you work with?

At ThingLogix we provide an IoT platform for our customers called Foundry that allows them to rapidly build the ‘software’ side of their IoT solution, whereby they make the data ‘actionable’ after collecting it from their IoT devices.  Our Foundry platform is a low-code serverless development platform that abstracts away the complexity of the underlying technologies from Amazon Web Services (AWS). We allow business and technical folks to configure their IoT software and in minutes build device digital twins, alerts/alarms, data transformations, workflows, reports and dashboards, machine learning algorithms, and much more.  Our customers can accomplish this without having to know how to code and integrate underlying AWS services such as AWS IoT, Kinesis, Cognito, DynamoDB, RDS, CloudWatch, API Gateway, and Lambda. We love Lambda and AWS Serverless technologies, because the technology allows almost infinite scalability, without all the headaches of infrastructure management. Of course, we are always excited to integrate new AWS features into our platform so our customers can take advantage of them with their IoT and other intelligent solutions they are building. For example, in the last year, we have integrated Artificial Intelligence using the AWS SageMaker service for machine learning. Our customers can now quickly select their devices and their device attributes and configure and test various machine learning algorithms based on the live and historical data they are capturing from their IoT devices. Even better, once they have an algorithm that they want to use, such as a predictive failure algorithm, they can then quickly incorporate that algorithm into workflows for processing new incoming data from their IoT devices. This way a prediction gets made every time new data arrives, and the prediction can drive tasks such as send an alert via text or email or open a support ticket in salesforce.com CRM for someone to check the device.  From the IoT device perspective, we work with all kinds of devices, protocols, and communication options…from endpoint sensors to gateways that aggregate data from tons of sensors, to 2G and 3G wireless, or SigFox and NB-IoT communications. Often it’s Bluetooth mesh to a wifi connected gateway for most industrial solutions, but consumer and field deployments have many options for configuration.

Read More: AiThority Interview Series With Scot Marcotte, Chief Technology Officer at Buck

The Industrial Internet of Things is understood to be a game-changer for heavy industries? How does ThingLogix intend to solutionize the problems of capital-intensive businesses?

Our customers in capital intensive businesses are most worried about their capital intensive devices not working and performing their duties required to do their job and produce their products.  They want to know if there is a problem with their expensive capital equipment before it shuts down their assembly line or negatively impacts a production run for a customer. A good example is our customer Toshiba who makes industrial size uninterruptible power supply (UPS) devices that are high capital purchases for their customers.  Now it may seem obvious, but UPS devices absolutely must work when power is lost.  The worst situation is if power is lost and you cut over to your UPS and the UPS fails.  So, we helped Toshiba outfit their latest fourth-generation UPS products with sensors that send data back to our Foundry serverless IoT platform for processing. Toshiba then monitors their deployed UPS devices on behalf of their customer using a web-based dashboard within Foundry that supports real-time alerts being sent to designated support personnel via text or email messages, web, and tickets in their CRM system. Toshiba can now give their customers peace-of-mind that the Toshiba UPS devices will be up and running when their Customer absolutely must have them operating.

Read More: AiThority Interview Series with Dean Abbott, Co-Founder And Chief Data Scientist at SmarterHQ

Predictive maintenance/analytics is one of the biggest boons of IoT. However, to get concrete results is time-consuming. How are companies like ThingLogix trying to solve this problem?

We believe a combination of rules-based logic capabilities combined with AI machine learning technologies is key to getting the best results with your predictive maintenance and analytics.  The reason is that some predictive maintenance items can be scripted out with standard formulas and rules-based analysis of data coming off your IoT devices when compared with historical information and “knowledge” or “rules” that subject matter experts (SME) are aware of and can help configure.  Foundry supports easy to configure formulas, rules definition, and workflow to configure your business logic for predictive maintenance rules that are based on SME knowledge. Now, the challenge is that many companies don’t have high-quality SME knowledge on what are the predictors of device failure or impairment. In those cases, we have partnered with AWS to implement AWS Machine Learning into our Foundry platform.  Once you have data coming into Foundry from your IoT devices, you can easily select the device attributes and run multiple machine learning (ML) models, such as Binary Classification Model, Multiclass Classification Model, and Regression Model.  You can then also use our Data Simulator to test and refine your models, and then put the models into production by incorporating them into your device data workflow using our Workflow Designer.  Unfortunately we don’t help you design your model, but we make it super simple to configure and test and deploy your models against all your IoT data being brought into Foundry, without you having to learn all the technical details of working with AWS services such as ML, DynamoDB, RDS, Kinesis, and much more.

 Read More: AiThority Interview Series with Daniel Clark, CEO at Brain.fm

Why does ThingLogix specifically choose/Promote AWS as its preferred partner for IoT solutions?

We have been a long term partner of AWS, starting from when we were part of 2Lemetry, which was sold to AWS and became their AWS IoT Core service.  ThingLogix was spun out and went on to build our IoT platform Foundry.  We are an Advanced Technology Partner, and IoT Competency Partner, and a Public Sector partner with AWS.  We strongly believe in the scalability and global cloud capabilities of AWS, which is critical for many of our customers who are deploying IoT solutions globally, not just in one country. We also are evangelists for Serverless computing, and we believe the AWS serverless stack is the best on the market and is a perfect fit for scaling IoT solutions like Foundry. Lastly, we value the emphasis AWS puts on security and privacy, so we leverage their services, including things like GovCloud to enable us to serve our public sector partners and customers.

How do you see the raging trend of including ‘AI in everything’ impacting businesses?
  • I think the biggest near term impact will be the shortage of Data Scientists to help companies evaluate, design, test and implement AI solutions.  Too many companies just do not have the right internal resources that are well versed in statistics, regression analysis, and many other mathematical techniques.  Companies are going to have to either invest in hiring and training this talent, or they are going to have to partner with companies that have a combination of technically skilled resources and industry solutions that can help augment their teams.
  • The second big impact on business is the need to improve their data.  We find that customers usually have data that is unreliable or is missing key information required to help build and refine their ML models. Often the customers know what data they need to collect in order to make predictive analytics work, but they often have not been collecting that data, so they need to build new collection techniques, and then collect data for months before they have sufficient amounts of data to begin looking at various AI models.
  • Finally, the big third trend is the sensor-enabling of everything.  If you want to really do predictive analytics you have to put sensors on your devices to gather the data needed to make predictions.  Sound simple, but it is not.  Sometimes this will require you to upgrade to new models of equipment, or other times you will have to purchase generic sensors and then customize the firmware and connections to work with your devices. The good news is that sensor prices are dropping radically and becoming commodity purchases in many industries.  What used to cost $500 a few years ago now costs $5 in many cases for sensors.  However, we find many customers are not even aware of how cheap sensors have become. Those companies who do take advantage of sensor-enabling their critical production, manufacturing or product devices will gain huge first-mover advantage in the future economy, allowing them to drive efficiency and offer improved customer experiences while reducing costs and improving their profit margins.

Read More: AiThority Interview Series with Kobi Marenko, CEO and Co-Founder at Arbe

The Cloud has substantially eased data management for enterprises? However, what do you think are the biggest challenges in Cloud deployments?
  • I would say the Cloud makes collecting data cheaper and easier, but the management of the data is still a challenge.  Many companies have very large amounts of data in the cloud, but they struggle to make it ‘actionable’ and ‘useful’.  That’s part of the big reason we built Foundry.   Foundry’s core business benefit is to allow customers to easily analyze, transform, and build business logic to make actionable all the data you are collecting from sensors and devices. We do this using a low-code model where business people can configure their solutions and only use their limited technical resources for customizations which they can build as small Lambda serverless functions.
  • Another big challenge with the cloud is finding qualified technical resources to work and manage your cloud solutions.  Finding someone who knows enough about AWS to build things from scratch is almost near impossible for many companies because the labor pool is so small for experienced people with AWS skills.  Colleges are not yet training on these skills at scale, so many companies are having to train-their-own resources, which takes a long time and often results in poorly architected solutions due to the inexperience of the teams.  We at ThingLogix saw this problem first-hand with our customers, and that is why we built Foundry to be an AWS best-practice based low-code platform that abstracts away the complexity of AWS and allows business or technical users to quickly configure their way to fully serverless solutions that take advantage of IoT, machine learning, and much more.
  • The last big challenge with the cloud is security. You cannot read the news today without seeing yet another company who exposed private or sensitive information from one of their cloud environments. This usually is due to the inexperience of the people managing the cloud environments. They don’t follow best practices for cloud security such as not using the AWS root account but instead using IAM role-based accounts, not using 2-factor authentication, not encrypting data in transit and at rest, and not using tools like CloudWatch to monitor your environment. None of these tools or techniques are difficult, but they all must be used in order to build that layered-defense you need to improve your security posture and limit the chance for exposure.  We built Foundry using AWS best practices, and we have participated in AWS architecture reviews to ensure we keep Foundry current with the latest security best practices. However, at the end of the day, our customers have to ultimately manage their security since AWS is a shared responsibility model, just like Azure or Google Cloud.  Leveraging platforms that already have security best practices built-in helps, but there are many operational, audit, review and ongoing maintenance items customers need to do and take seriously to stay on top of security in the cloud

Read More: AiThority Interview Series With Scot Marcotte, Chief Technology Officer at Buck

Being the head of operations, how important is it for you to keep in touch with the technological evolution of your company’s offerings?

Extremely important.  We are a small company, so we all wear multiple hats.  I still get involved with key customer sales activities and ongoing customer delivery. You cannot effectively help sales, or ensure delivery if you don’t know your products, the capabilities, and how to maximize the benefits for your customers. Thankfully we have our CTO and his team who walks us through all the new features and benefits every 2 weeks on a web conference. In addition, I write much of the product marketing content for our website, and assist in writing content for our subscriber emails, and our blogs. I collaborate with the tech teams to make sure I’m adequately representing the technology while conveying the benefits in business terms that our customers understand.

Does ThingLogix intend to infuse Machine Learning in the Business Applications it makes from Java?

We are a node.js serverless shop and we strongly believe in serverless computing. The good news is Lambda functions as a service can be built in Java, but it also supports 7+ other languages, so we have chosen node.js because that is where the market is heading and where our customers also are strategically moving.  Of course, as mentioned previously, we take advantage of the AWS machine learning services and we already infuse those in the Foundry platform in a way that abstracts away much of the complexity from building, testing, deploying and utilizing machine learning models.  This allows our customers to infuse machine learning rapidly into their business applications, which allows exponential increases in speed of deployment of ML solutions to the market.

Read More: AiThority Interview Series With Tom Livne, CEO & Founder, Verbit

What’s the most exciting part about leading an AI business in the technological ecosystem?

The most exciting part of leading and AI business in the technological ecosystem is how fast we can bring benefits to customers. Everyone wants to move faster, and stay ahead of their competition, or launch innovative new business models to differentiate themselves or create new markets.  Thankfully the technology has finally evolved to be able to keep pace and help accelerate, rather than slowing, new business solution development and deployment. Low code platforms like Foundry, built on infinitely scalable serverless technology from AWS, when combined with the ability to quickly gather, analyse, interpret, and ultimately turn data into competitive AI supported business logic that differentiates your product…that’s a huge win for our customers, and it gets them super excited, as it does us at ThingLogix.

Which events and webinars would you suggest to our readers as being the best in grasping information on emerging technologies?

There are a bunch of great free training webinars available from leaders in this space including AWS, salesforce.com Trailhead, and Google.  You also can find very low-cost webinars, training, and certification from Udemy.  Our own employees utilize these resources consistently to help themselves stay on top of technological trends. Also, there are great websites dedicated to AI such as AIThority which covers the business implementation of AI services in Sales, HR, and Marketing, while also helping keep subscribers abreast of things happening across the entire AI ecosystem.

Read More: AiThority Interview Series with Kimberly Powell – Vice President, Healthcare at NVIDIA

What geographical regions (The Americas, EMEA, APAC) does ThingLogix cater to?

We primarily focus in the U.S.A., Europe, and the Middle East. We are headquartered outside San Francisco, but we have sales and delivery resources around the world and a GM in Europe and a GM in Dubai.

What are your thoughts on the misuse of Artificial Intelligence?

The challenge with using data to make decisions is that garbage data results in garbage decisions, which can create enormous problems for people and businesses. This is not a new phenomenon with AI. Rather AI technologies can, unfortunately, accelerate making bad decisions when you don’t properly train your models and build your algorithms. Ultimately companies utilizing AI need to adopt the principles of scientific methods which has been around since the 17th century. Companies need to develop AI solutions while also applying rigorous skepticism about what is observed, given the cognitive assumptions that can either distorted or misinform the interpretation of the results, thereby creating problems with the implementation of AI.  We certainly see the need to make sure that AI is rigorously tested to ensure you don’t have inadvertent misuse in critical areas like healthcare, government policy, finance, and many other areas.

Read More: AiThority Interview Series with Anil Kaul, CEO at Absolutdata

Where do you see AI/Machine learning and other smart technologies heading beyond 2020?

Technology usually follows both the hype cycle and we are likely heading into the trough of disillusionment for AI.  Many companies are trying to apply AI in areas where they are not likely to succeed due to many factors, and this will likely cause a pull-back in excitement.  However, technology trends are so fast that these long hype cycles of the past are being replaced with very compressed hype cycles of today. This means we will likely move rapidly past the trough of disillusionment and into the slope of enlightenment and plateau of productivity.  Technology really becomes powerful in the production phase, when the complexity of the technology is abstracted away and the technology becomes easy-to-use and integrate, thereby allowing rapid adoption and accelerated deployment across many industries.  We fully expect this to happen in 2020-2022 and we expect to be part of the solution to bring easier-to-use AI to business customers.

What start-ups are you keenly following?
  • We are lucky enough to have a number of startups as customers, and we keenly follow their progression and expansion.  Xenon AI out of LA has a unique IoT solution for managing huge volumes of assets with real-time data analysis for equipment status and control.  O2 Concepts has an innovative IoT connected portable oxygen concentrator that allows their customers the freedom of safe, reliable and limitless oxygen.  SolarNow out of Uganda that sells and finances IoT connected solar home systems for East Africa to allow families to have quality, sustainable energy solutions at affordable prices.  Saferide out of the USA, provides an IoT connected set of devices and software to protect consumers and business drivers from distracted driving.
  • We also follow innovative hardware and connectivity startups such as eseye who help companies connect their IoT devices to any network in any country, with full security.
  • Lastly, we follow innovative Edge Computing companies like edgeworx, who offer an open source based edge computing platform to allow you to deploy microservices to edge computing devices.
A piece of professional advice that changed your life?

The number one thing that allowed me to succeed was focusing on people and building great teams.  The best advice I ever received was to focus on building my successor. I initially did not understand why I should focus on this aspect of career development. However, it was explained to me that there is an interesting paradox that flummoxes many aspiring leaders. To be considered for management early in your career you have to be one of the most knowledgeable people on your team in your department.  However, as you move up in the management hierarchy there is a distinct change in how you are perceived to be ready for further promotions.  Specifically, if you continue to be the most knowledgeable person in your role as manager, this will send clear signals to senior management that you are (a) unable to delegate effectively, and (b) you are not ready for promotion because you are too valuable in your current role because you have not yet built your successor.  I have often heard a phrase like, “Oh, we can’t promote <name>, he’s too valuable in his current role.”  That is a signal that you have made yourself indispensable in your current role, usually by not delegating and building your successor, so you will not be considered for promotion and you will fall off the management track.  Now that I am older and wiser, I try and tell managers that another reason to effectively delegate and build your successor is that doing so will actually free up your own personal time so that you can take on stretch assignments or additional duties and responsibilities.  Senior management likes those leaders that can take on stretch assignments while still effectively managing their teams.  This shows agility, the ability to juggle multiple balls, and, of course, it shows that you have built a strong team that can operate without you having to be involved in the details.  Senior management is always on the lookout for leaders who can build leaders.  Strangely, very few managers focus on this key aspect of management for fear that someone they manage will become knowledgeable enough to perform their job.  That is actually a good thing.

Read More: AiThority Interview Series with Michael Scharff, Chief Executive Officer at Evolv

Tag the one person in the industry whose answers to these questions you would love to read.

That is easy…Werner Hans Peter Vogels, the VP & TO at Amazon Web Services.

Thank you, Eric! That was fun and hope to see you back on AiThority soon.

Eric Dirst is the Chief Operating Officer of ThingLogix and is responsible for day-to-day operations across Sales and Marketing, Customer Success and our Partner Program. With over three decades of technology and operational executive leadership experience spanning technology consulting, management consulting, startup leadership, CIO for Fortune 500/1000 companies, and President/COO for a Fortune 1000 company. Eric is also proud to serve on the Boards of Carrington College and the Brent Sopel Foundation, as well as Advisory Boards for DocuSign and IT Central Station.

ThingLogix was founded in 2014 and was originally the services group of IoT platform provider 2lemetry. In 2015, AWS acquired 2lemetry, which became the basis for AWS IoT. ThingLogix has maintained a strong relationship with AWS, architecting Foundry to orchestrate all of the AWS services necessary for a complete IoT solution, and also certifying as an AWS Advanced Technology Partner.

Leave A Reply

Your email address will not be published.