To stay ahead in tech, companies must remain in a constant state of evolution. Whether it’s a move to the cloud to speed up software deployments or the implementation of a cutting-edge technology like artificial intelligence to improve automation, the companies leading the pack in their respective industries are those that aren’t afraid to embrace change.
We recently caught up with six NYC companies to get a behind-the-scenes look at how their technology has evolved since their early days and what they’re most excited for in the months and years to come.
Since its founding in 2011, process mining technology company Celonis has seen its product move from an on-premise solution to a cloud-based platform, steadily defining a new category of tech all the while. Solutions Engineer Caitlin Schultz explained how their technology has evolved since the early days and how they’re leveraging AI and automation to continue pushing the industry forward.
How has your technology evolved since the early days of the business?
When Celonis first started, the on-premise-based product helped analysts understand how processes drove business efficiency. We knew event logs could be leveraged to create a data-driven view of a process instantly, rather than a subjective perspective cobbled together from interviews.
Today, we offer a containerized, cloud-based platform that helps enterprise companies become superfluid. Our product suite helps thousands of users get timely information, be more efficient and make smarter decisions. We’ve developed products like the Action Engine so users across any skill set can anticipate process friction and easily improve the business outcome. At the same time, we’ve reduced the time-to-value and expanded our product’s reach with our App Store, a set of hundreds of pre-built analyses, based on 2,000 implementations and the broad knowledge of our partners.
We don’t want to just tell people that there’s an issue; we want to recommend the next best action for them to take — and enable them to take it with a single click.”
What will this next evolution of your tech look like? Are there any new technologies you're eyeing for the future?
We don’t want to give away too much, but we’re expanding our AI and process automation capabilities within the Action Engine and across the platform. That’s really driving this Enterprise Performance Acceleration category that we’re defining. We don’t want to just tell people that there’s an issue; we want to recommend the next best action for them to take — and enable them to take it with a single click. Along with our investment in AI, we’re making process mining accessible to anyone, offering the first free cloud-based process mining tool in the market, in our recently released edition, Snap.
Document management has, until recent years, been a largely manual process — a necessary but rather time-consuming one. Ocrolus stepped onto the scene to change that in 2014, and five years later, they continue to keep an eye toward innovation. VP of Engineering Zoheb Sait explained how his team is leveraging increasingly sophisticated technology to make document automation faster, more accurate and more secure.
How has your technology evolved since the early days of the business?
Ocrolus captures and analyzes submitted documents, such as bank statements, tax forms and paystubs, turning unstructured and semi-structured assets into machine-readable data. At the beginning of the company, Ocrolus used standardized OCR to extract data from financial documents. As Ocrolus grew, the platform combined AI-based computer vision with crowdsourced review to extract and analyze data on financial documents with more than 99-percent accuracy. Today, Ocrolus combines artificial intelligence, machine learning and hyper-efficient human quality control in a processing engine that performs all stages of document review in an end-to-end process — from classification and extraction to validation and analysis.
The next evolution of our tech will weave advanced technologies together to take document analysis to a higher level.”
What will this next evolution of your tech look like? Are there any new technologies you're eyeing for the future?
The next evolution of our tech will weave advanced technologies together to take document analysis to a higher level. Deep neural networks will enable Ocrolus to understand unknown documents by comparing them with thousands of other examples with precision point analysis. This is important in industries that mandate the analysis of hundreds of diverse documents, such as the mortgage sector, which requires our processing engine to ingest and comprehend unknown inputs that it has never encountered before. Ocrolus is also adopting new anti-fraud detection by using file forensics, alongside expert review and network intelligence, to stop bad actors in the loan application process.
Marketing today looks a whole lot different than it did when Movable Ink first launched in 2010, and while the personalization of marketing content — particularly email — has always been the company’s objective, recent advancements in AI and big data have allowed the company to push the boundaries of personalization even further. Senior Technical Writer Gerrie Cho shared how the company is leveraging these rapidly evolving technologies to elevate creative visual experiences for their customers.
How has your technology evolved since the early days of the business?
Movable has always been about creating amazing visual experiences, and over the years, data has become even more integral to what we do. What started out as simple applications of data to do things like image personalization and countdown clocks has transformed into huge advances both in how we access data across a multitude of internal and external sources and how we use the data to activate creative at the moment a customer engages. Movable Ink has also been focused on creating easy-to-use tools to help marketers automate the development of creative to eliminate the inherent bottlenecks with creating visually engaging one-on-one experiences.
Now, in a few steps, you can connect to APIs or other data sources to infuse marketing creative with meaningful, maybe even inspiring, moments that are uniquely special to every person. This means a bargain hunter may view curated product recommendations and discounts (along with a real-time inventory of items in stock), while someone who’s more mission-minded may see stories about a company’s charitable work along with opportunities to get involved based on their current location.
...We’re excited about the tremendous opportunities that AI can open up for us and our clients.”
What will this next evolution of your tech look like? Are there any new technologies you’re eyeing for the future?
Plugging in data is only the first step to creating content. You also need a lot of creative to make the data come to life. For many companies, generating this personalized content can be a bottleneck. With the help of artificial intelligence, we’re exploring ways to solve this problem by combining customer interests with marketers’ goals to predict the best-performing creatives. These creative variations can then be tested against each other to determine what resonates with particular customer segments. AI has the opportunity to produce incredibly powerful insights and results, and we’re excited about the tremendous opportunities that AI can open up for us and our clients.
Hopper didn’t initially launch into the travel tech market with its mobile flight predictions app, but that’s certainly how it made a name for itself. Paring its millions of historical data points with artificial intelligence, the company quickly realized the power big data could have in transforming the way we shop for flights. Principal Software Engineer Mariana Baca shed some light on Hopper’s transition to a mobile-only application and where the company is headed next.
How has your technology evolved since the early days of the business?
Hopper’s first iteration was a travel website. As the business progressed, however, we realized that by leveraging our millions of historical price points and AI we could help consumers get the best deals on flights by predicting when flight prices would go up or down and the ideal time to purchase. At this point, Hopper pivoted to become a mobile-only flight predictions app. Having a mobile-only platform allowed us to use our big data to maintain an ongoing conversation with Hopper customers via push notifications, informing them of the ideal time to purchase a flight. Our technology stack is now built on Scala microservices, with native iOS and Android front ends, which allows us to rapidly iterate the product and quickly add more value for our customers.
We are continuously looking into new ways we can ensure the best prices for our customers.”
What will this next evolution of your tech look like? Are there any new technologies you're eyeing for the future?
We are continuously looking into new ways we can ensure the best prices for our customers. One technology we are testing right now involves using our prediction technology to provide a new pricing capability for our customers, essentially allowing them to freeze the price of a flight so they have extra time to decide before purchasing the ticket.
We will soon be embarking on a project to migrate our physical infrastructure and data to the cloud from physical data centers. Having our infrastructure on a public cloud platform will allow us to continue to grow and scale at the rate we need to, but will certainly create some interesting problems to solve as we adjust to the new technology landscape.
In the future, we are looking into how we can improve our prediction models when customers want to buy tickets for larger groups so we can continue to improve our flight availability predictions and provide a better experience for our customers.
We also want to expand into new markets, including Latin America. Many low-cost carriers in that part of the world are not plugged into traditional travel inventory providers, which means one of our technology challenges will be supporting different connections with travel partners in the app, including direct connections and relationships and proprietary APIs.
Hotel digital marketing and technology company NextGuest Digital got its start in the content space, helping hoteliers elevate and manage their digital presence with a simple CMS platform. The company has since expanded its tech offerings to encompass marketing, CRM and a whole lot more, and according to Executive Vice President of Strategy Margaret Mastrogiacomo, that’s only the beginning.
How has your technology evolved since the early days of the business?
As one of the first companies to enter the hotel digital marketing and technology space, we recognized the importance of hotels having fresh and relevant content to attract and engage website visitors and turn them into bookers. To accomplish this, we created the hotel industry's first content management system that gave hoteliers the power to own their digital presence. What started as a simple toolkit for website text and imagery has now evolved into a website technology platform that includes personalization, marketing automation and more that was built for hoteliers, by hoteliers.
Looking into the future, our goal is to expand our martech to address other touchpoints in a traveler’s path to purchase...”
What will this next evolution of your tech look like? Are there any new technologies you're eyeing for the future?
Fast-forward to today, we have expanded our martech to include not only a CMS but also a CRM platform. The integration of these two platforms allows hoteliers the opportunity to not only acquire new guests but also engage and retain past guests. However, it doesn't stop there. Looking into the future, our goal is to expand our martech to address other touchpoints in a traveler’s path to purchase and have a fully integrated end-to-end solution that helps hoteliers engage with their future, current and past guests. Our mission is to be the best and most innovative hotel technology platform in the industry.
The rise of cloud technology has opened up a whole new world of possibilities for tech companies of all stripes. For Stocktwits, in particular, transitioning away from physical servers to the cloud has transformed the way they develop and deploy fixes and features for its platform, a social network geared toward investors and traders. Chief Technology Officer Anthony DiMarco shared just how substantial that impact has been and how they plan to double-down on a serverless approach.
How has your technology evolved since the early days of the business?
Over the last five or six years, we've made the transition from an antiquated way of thinking, where we deployed code to a fixed number of physical servers in a co-location facility to running our infrastructure completely in the cloud. We were deploying a single, monolithic piece of software by hand every two to three weeks. Carefully. We can now deploy whenever we have a new bug fix or feature that needs to go out, even during production load in the middle of the day, often several times a day.
Our first step was a lift-and-shift of our entire production infrastructure to AWS, where we started using the 'immutable server pattern' — packaging each release as an AMI with a repeatable process and describing our infrastructure with code in the form of CloudFormation templates. This was a relatively small conceptual jump that brought tangible benefits. Suddenly we could scale elastically and roll back with ease.
From there we made the transition into EKS, the managed Kubernetes cluster offering from AWS. It provides similar deployment patterns with Docker containers instead of disk images, but with a much faster turnaround time for deploys since we don't have to wait for physical hardware to spin up.
The next evolution of our tech stack is fully serverless.”
What will this next evolution of your tech look like? Are there any new technologies you're eyeing for the future?
The next evolution of our tech stack is fully serverless. All of our new development makes heavy use of AWS Lambda and API Gateway, and instead of building deployment images and caring about scaling them, we simply provide function code to AWS Lambda, which manages the deployment and scaling for us. We no longer have to think about or manage servers at all, which frees us up to focus on feature development instead of infrastructure maintenance and planning.
