Dror Sgan-cohen is the head of delivery and IT for Precognize. He shared some of his experiences onboarding SAM GUARD® customers to use the solution effectively. 

Can you explain what is involved in serving as head of delivery?

Head of delivery means that I’m responsible for implementing and managing each customer’s SAM GUARD® project. The delivery team installs SAM GUARD®, configures the system, and prepares it for customer use. After initial onboarding, we monitor the system and the server that’s hosting it, to make sure that it’s always available and working correctly. We carry out upgrades and fixes, and deliver customer support. When I joined the company, my title was Customer Success Director, and although the title has changed, the role is still similar. 

I carry out some pre-sales work too. Often technical questions come up during the sales process, and if the sales team needs a more technical response, they turn to me.

What does a typical day look like for you?

There’s no such thing as a typical day! My day is divided between planned tasks on my to-do list, like delivering upgrades and updates, replying to customer requests, that kind of thing, and urgent issues that suddenly come up like monitoring alerts. Since I never know what will come into my inbox, each day is different.

What is your professional background?

I’m a computer engineer by profession. I graduated from the Technion (Israel Institute of Technology) in Haifa, and then worked for the telecom industry at Amdocs for 15 years in development, delivery, support, and operations. I worked my way up from developer to team lead, senior team lead, then project manager, and I relocated for four years to the US and Canada to support Amdocs’ customers there. 

After I left Amdocs, I worked as a freelancer mainly for E-plus, which is Germany’s third-largest mobile operator. And more than five years ago I joined Precognize. 

My computer engineering background means that I can handle all the technical tasks I need to do, and I still like to get my hands dirty with the tech. I try to keep up with technology so that I can do things myself when I can, instead of assigning them to other members of my team. I’m interested in how things work, what the logic is behind them, that kind of thing. 

What made you decide to work at Precognize?

Until I came to Precognize, I’d only worked in mobile and telecom companies. Precognize was in a field that was totally new to me, and I was ready for something different. I was excited to learn more about machine learning and artificial intelligence since this is where the world is going, and I was particularly interested to learn about what SAM GUARD® is doing, what it’s contribution would be to the field, and the kind of service it is delivering.

Can you walk us through the process of onboarding a new customer?

The first step is to set up the server where SAM GUARD® will be installed. If we’re hosting the product, our IT (which I also manage) establishes a new virtual machine to host SAM GUARD® for the new customer. If we’re installing it on premise, I coordinate setup with the relevant people on the customer side, get details of the server and remote access. 

Then the team installs the latest version of SAM GUARD® onto the server, and carries out the default basic configurations needed for it to work. Next, we create user accounts so that the customer can login and use the application. 

After that, there are two main steps which need to take place before SAM GUARD® is up and running. One is to build the plant model within the application. Sometimes it’s done by someone from the Precognize Analytical Monitoring Service (AMS) group, and sometimes by someone at the customer end. We have tools for the customer to make this easy for them to do on their own. 

The other important task is to upload a year of historical data from all the plant’s tags, which is an enormous amount of data. SAM GUARD® has connectors to most of the commonly-used data sources, so if the customer uses one of those, we can just connect SAM GUARD® and load the data. But if not, they provide us with files, and we’ll load everything into the database. 

Once the historical data is loaded and the plant model is built, we’re ready to start SAM GUARD® running online with the continuous data feed. Either the customer connects their data feed, or they send us data files once every hour or so, and we add them to the server. From that point on, the system is live. 

In your experience, how should process plants prepare for a successful implementation of SAM GUARD?

The most important thing they can do is to gain management support and have them encourage the plant team to adopt the solution. Because process engineers are busy all the time, so they’ll often say “I don’t have time for another tool.” Management needs to convince them about the value of SAM GUARD®. 

Another big step they can take is to set up an AMS team. They are the first line looking at SAM GUARD® alerts, and they then talk to the process engineers about the most important alerts or periodic reports. Some plants use our AMS team, but we also have customers that set up their own AMS teams, working together with us. They have a single centralized AMS group that deals with alerts and reports for all the company’s plants, and then they get into contact with the process engineer for the relevant plant when something comes up. 

What do you like best about your work?

The best part is the people I work with. Even though we’re part of the large SAMSON Group, Precognize itself is small and we all know each other. I really enjoy my daily work, especially since even after so many years, I’m still learning new things. And lastly, I really like being in contact with the customers, supporting them, and getting positive feedback when everything goes right. I get satisfaction from their satisfaction.

What do you find most challenging about your work?

The most challenging times are when there’s a lot of pressure to resolve critical issues fast, or when there are a lot of tasks to deal with at once. I want to make sure to give our customers the very best.

Can you share a story about a particularly interesting delivery you did?

Long before COVID-19, in one of my first years with Precognize, I had to go to Brazil to manage the implementation for a new customer, a paper plant. First I had to fly to Madrid, then to Sao Paulo in Brazil, then to another city in Brazil, and then there was another 3+ hours of driving. 

I was met by a local agent, and he drove me through the jungle, across a plain with nothing at all in sight, and then across a desert. Slowly a small building appeared in the distance, it looked just like something out of a movie, and he stopped by this one-floor building and said “this is our hotel.” I said “Are you sure?” and yes, he was sure. It was just a few rooms and an office, that was the hotel. 

Well, it was Sunday evening, and we were hungry, so we drove to the nearest city – they called it a city but it was really a village – to try to get something to eat. But everything was closed. Eventually someone directed us to a nearby restaurant, which turned out to be our hotel! It didn’t actually have a restaurant, but the man in the office asked us what we’d like to eat, and he went into the kitchen and made us some food. 

The next day, we went to the plant, which is enormous. It’s one of the largest raw paper product manufacturers in the world. But it’s totally cut off, there’s not even any cell phone service at the plant. I was there for two weeks setting up SAM GUARD® for them and training them on how to use it. The implementation went smoothly, but the travel and the working conditions made it a real adventure. I had no clue what I was getting myself into before I went there.

What changes do you anticipate over the next 5 years, in the world of predictive maintenance and predictive monitoring?

Process plants are going to adopt more digital tools that get more value from their data. This is the direction the world is moving in. 

Right now, plants are measuring and gathering data from tags and sensors, but they aren’t doing much with that data. Without advanced technology, the control room can only handle a few dozen tags, so they just look at the most critical ones. But predictive monitoring and predictive maintenance systems like SAM GUARD® that use AI and ML can cope with all the tags, all the time, 24/7. 

I expect that more plants will understand that they need a tool or system like SAM GUARD® that can analyze their massive data volumes and pick the most relevant items to be presented to the professional team for action.