I recently spoke to Varun Goswami, Vice President of Product Management at Newgen Software. Varun joined Newgen in 1997 as a programmer and made the switch to Product Management during his career there. He was promoted to his current position in December 2022. His role is to conceptualise, design, and build the products. He also owns the messaging and the GTM part of the product for Newgen Software’s core products.
That core product is NewgenONE, a low-code platform that enables customers to automate their digital transformation journeys. NewgenONE has a comprehensive set of tools, including process modelling, interface building, data modelling, real-time reporting, as well as content and communication capabilities. Varun believes that this diversity gives the product a competitive edge over other low-code and other point solutions in the market.
Using NewgenONE, Newgen can deliver an end-to-end digital transformation journey for customers. It uses customer-facing portals and mobile apps, automating processes from the middle and back office and integrating with legacy and core systems.
I asked Varun about Newgen Software.
“Newgen is a 30 year old software product company based out of India. Last year, we did $125 million in revenue. We operate in 72 countries, and we have subsidiaries in six locations across the globe. Our primary focus is on low-code application development, content management and communication management. We have around 3,000 employees. Out of these, around 570 employees are in core R&D. We have applied for 44 patents, of which 23 have already been approved. We help our customers accelerate their digital transformation journeys through our innovative products.
“Our revenue is growing almost 20 to 25% — year on year. We’ve been a profitable company throughout. We are a publicly listed company with our shares traded on both the major indexes in India, which is the BSE, Bombay Stock Exchange and NSE, the national stock exchange.”
What is the target market?
“The target market is the whole low-code automation and communication automation market, which is about a 26 billion TAM. It’s spread across different geographies.
“We primarily target the Fortune 2,000 companies. We target large and mid-level enterprises. Almost 70% of our revenue comes from financial services companies, banks, and insurance companies. The rest comes from government, healthcare, BPO shared services and other domains. The products that we build are very horizontal, so the same product can be applied for automating transformation journeys across different industries.
What’s the product architecture?
“We have four layers: the presentation layer, the business logic, database. We additionally have a fourth layer, which is the storage layer, because we deal with a lot of content.
“For example, if it’s a PDF file, we have to store the metadata in the database, but the PDF file content cannot be put in the database. So, we have a storage layer where that content goes.
“On the database layer, we support RDBMS, like Postgres, MSSQL, and Oracle. On the business logic layer, we have support for Java EE, and we recently migrated into micro-services. At the presentation layer, we use the React framework, and on the storage management layer, that’s a custom layer to store the entire document data.”
What is the release schedule for NewgenONE?
“We have a major, minor, a service pack and patch release. A major release is when we add some major new functionality, we are currently in the process of adding generative AI capability. That typically happens once every 12 to 18 months.
“Minor releases are where we are enhancing our product functionality, which typically happens every six to eight months. Service packs are a collection of patches, and we patch whenever there is an issue. Then, every few months, we combine all of those patches and release the service pack. If a customer has any issue, they can apply the patch, or they can wait for the next service pack to come, which happens every three months.”
What’s on the product roadmap?
“We have lots of things going on. Video content has become very popular in the last two or three years. We’ve built out the functionalities for video content storage. Once you’ve stored videos, how do you search for them? Video summarization, video analytics, and video streaming are other areas that we’ve worked on.
“Architecture, with the microservices release we’ve recently done. There’s still some way to go. We still use some enterprise Java for the designing part, probably that’s something that we are going to target. We have a low code transformation journey builder. We are looking at how inbound and outbound communications can be combined. A couple of years back, we bought a low-code data science company called Number Theory. So we are in the process of integrating that into our core products, and of course, there is generative AI.”
What have you achieved this year,
“We had a major release called NewgenONE 2023.1 this year. Last year, we had three separate products called iBPS, which was for your local application development. We had OnmiDocs, which was content management, and we had OmniOMS, which was communication management. We have built an Automation Studio, in which you have a low code application design, content services, communication services, and artificial intelligence.
“This year, it was one of the biggest releases that we’ve done in our history. We have a single platform where you can apply intelligence, build any application, build the interfaces and the business rules, can define what content that application is going to carry. It will seamlessly move from one person to another and (support) any outbound communication that you want to send, whether it’s for insurance, your contracts, life policies, or reminders. You have digitally signed receipts. All of that can be done through a single platform.”
What do you hope to achieve by the end of 2024?
“Generative AI. We’ve got a generative AI product called NewgenONE Marvin. The product release is done, and the marketing release will happen pretty soon. We’ve just scratched the surface of generative AI capabilities. It’s very exciting, and we already have capabilities for searching documents using natural language queries.
“We have capabilities for creating hyper-personalised communications at runtime and even building entire data models or application journeys through generative AI. We are looking at definitely enhancing that NewgenONE portfolio. Beyond that, we have a whole product roadmap, in terms of content in terms of communication, as well as in terms of enhancing our low code capabilities.”
What’s the future hold around generative AI, looking one year in advance, and then five years into the future?
“We are looking at two different aspects. One is in terms of the capability. The second is in terms of privacy and the ability to handle those things correctly without hallucinating. It is why we built the NewgenONE Marvin layer because we did not want our customers to take a course on prompt engineering and then come back and try to figure out, how do I ask this question. We’ve created the interfaces, and we’ve allowed them to mention what they want to build. Then we do the prompt engineering, and we get the data, parse it and paste it on the interface. We have taken a very different approach to generative AI.
“Five years down the line, I won’t be able to give much input. What we are planning to do is we want to make sure that our customers benefit from the capabilities that any of these large language models provide.
“One of the areas that I can talk about is the image generation part. I talked about how we can build processes faster. We can build communications, but what if a communication wants to have a banner? I’m launching a new type of home loan, which is 1.5% better than anybody else, and I want an image or a banner to be generated for that.
“Those are capabilities that are still in the nascent stages. We’ve experimented, and they are still not that good. Text is very well handled today. In the next year, we are going to focus on text to get a lot of capabilities out of the door. But images, video and other multimedia content will take at least two or three more years to mature.”
What different kinds of technology partners do you have?
“We’ve got technology partners, where we end up consuming the technologies. Red Hat and Oracle fall under that category. There are certain partners like Amazon and Microsoft where we host our SaaS platform on their cloud service. Then we have ISV technology partners, for example, SAP or Finacle. We do a lot of image enablement for them. Wherever there is a need for a document and the data to be seen together and processed, or a video, for example.”
“SAP is great at managing transactional data, in terms of invoice data or HR data. Wherever they have to store documents, if I have to store the entire document, the invoice as a PDF, JPEG file, or TIFF file, they struggle. We enable that, we expose our API, and they consume our API to store the invoice data. Whenever a user needs to approve the invoice, they call our API to fetch the invoice, and we store, retrieve, search and view for SAP both for R3 and S/HANA.
“We do a lot of integration with Duck Creek, which is a major insurtech company. So whenever there is an insurance claim in which somebody needs to review a video, add annotations or look at the documents and then finalise whether the claim has to be passed or not, we build those front-end interfaces and those orchestration layers.”
The book question
What was the latest book you read?
“I’m currently reading on audible Yuval Noah Harari’s “21 Lessons for the 21st Century” (Amazon Aus, UK, US). The last book that I read was Ikigai by Héctor García and Francesc Miralles (Author), Amazon (Aus, UK, US). That’s the whole concept of what you love, what pays, and your reason for working or your reason for being. It’s a nice book and a light read.