It’s taken people some time, but most of society is finally doing a collective 180-degree flip on its once guarded view of artificial intelligence in favor of being open to the idea that it can help humankind in a myriad of positive ways.
That’s why Houston-based software development company Allganize recently secured $20 million in series B funding to propel its industry-leading AI answer bot, Alli.
The All-in-One LLM Enabler Platform, which intelligently responds to customers through natural chat conversation flows, while also enabling the automation of up to 80 percent of support tickets, allows customers to get serviced faster while improving employee productivity.
With the latest $20 million investment by InterVest and Murex Partners, that brings the total funding into Allganize to $35 million. That investor confidence will ultimately help catapult the company’s AI solutions to the next level and help target its planned Japanese Stock Exchange listing by 2025.
“We will lead the expansion of corporate-specific LLM app markets and accelerate the distribution of enterprise automation AI in USA, Korea, and Japan,” says Changsu Lee, CEO of Allganize. “We are dedicated to empowering companies to develop custom LLM applications, enabling practical tasks execution and work automation.”
From programmer to prototype
Allganize’s plans to go public in 2025 stems from Lee’s early ties to Japan. While Lee is originally from Korea, he got his start in Japan, where he was able to secure his first job as a programmer at a gaming company and years later the first investment for the first company he started, ABLAR.
“I'm originally from Korea and when I got a scholarship from the Japanese government, and I went to the research laboratory of Tokyo Institute of Technology, I was not able to speak Japanese at all,” Lee tells InnovationMap. “But that one year actually changed my life. I could speak Japanese while I was there, and I was able to learn a lot of Japanese cultures, and there in the social system, I became a big fan of Japan, and then after that, so I was really looking for an opportunity to do business with Japan, or in Japan.
“I was originally trying to start a company in Japan, and I started to work as a programmer in a gaming company, and at night time, and also the weekends, I was actually building a product to start a company," he continues. "So back then, that product name was ‘Search for You.’ It was every single person's search history, the trees, then if we can find somebody else who's already advanced, but pretty similar search tree, then I was thinking, we can probably, give better suggestions or the recommendations. This is future knowledge that you could probably expect to learn, and if we can have a million users, then we can leverage, those search histories from Google."
Changsu Lee is the CEO of Allganize. Photo via LinkedIn
From there, Lee built the prototype, while making the bulk of his decisions from Japan. Later, while back in Korea, he expanded his business into Japan and met the biggest VC of Japan at an event in Korea and gave a quick pitch of his company’s service. Ultimately, he got the funding and hired Allganize co-founder Yasuo Sato, who he’s been working with for more than 11 years now.
For Lee, it all comes back to where his first idea and mission was believed in and supported, and that place was Japan. Not surprisingly, the bulk of Lee’s customers are in Japan, and it was there that he began his AI journey, by cutting his teeth in machine learning.
“After changing the company name to 5Rocks, we were providing analytics and marketing automation solutions for mobile game companies,” says Lee. “So even back then, we were actually using machine learning, but it was not deep learning. We were using machine learning to predict every single gamer's remaining lifetime in the game. For example, how many days or how many months is this gamer going to stay in the game and how much money are they going to spend in the game for their game items?
“We were predicting this by using machine learning," he continues. "And after we were acquired by the mobile advertising company, I worked as the SVP of the platform there, and I actually got the opportunity to learn the deep learning.
If it’s possible to dig deeper into deep learning, that’s exactly what Lee did as the senior vice president of Tapjoy. Selling his first company and staying onboard in a much more limited role allowed Lee to have a sort of paid internship or technological rotation program into AI and deep learning, which he believed was going to completely change everything and, as a result, was to be the major foundational key to his next venture, Allganize.
Japanese IPO on the horizon
Fast forward to 2023 and the $35 million in investments, Lee and his team are aiming to expand the company’s existing customer base of over 200 enterprises and an IPO on the Tokyo Stock Exchange in 2025.
Take a quick look around and roll call currently known AI platforms. Google Duet, Microsoft Co-Pilot and, of course, ChatGPT quickly come to mind, right?
Spoiler alert: those are all descendants of Allganize’s Alli All-in-One LLM Enabler Platform, which it continues to enhance. Still, Lee and his team still find themselves having to reassure that AI is not the boogeyman movies like The Terminator or Ex Machina have made it out to be.
“I think the most important thing is in how we can leverage and utilize these AI tools,” says Lee. “And because, at the end of the day, it's still a tool, AI is great. It's really powerful. But it's still the tool for helping humans. For example, for white-collar knowledge workers, productivity is completely different when they are using the AI. AI is also more knowledge-focused, and you can answer all the questions that you have.”
Simply put, AI can add not only add to efficiency and productivity, but it can also answer questions before anyone can think to ask them and fill in the blanks to missing information that one didn’t know was missing or could think was missing.
“As powerful as that sounds, AI still can't replace people because people still have to direct it and guide it to where it needs to go,” says Lee. “And because AI's capability is becoming more and more powerful, it’s important for us to learn how to safely train the AI and how to make the good guidelines that we're using for AI.”
With Allganize focused on large language models, they assist businesses to leverage AI to enhance their employees’ knowledge recall and operational efficiency.
“I'm always saying the LLM model is just an engine of the car,” says Lee. “People are buying cars, people are not buying engines. Of course, the engine is important for the car's performance, but people are actually buying cars, not the engines. So, we are really focusing on what kind of application you really want to do in your organization by using LLM.
"And again, our ultimate focus is all the applications because this AI should be used in the daily life of the enterprise, which is the workflow," he continues. “We call them the LLM app. At the moment, the biggest LLM app is called the cognitive search. So when an employee or the customer is asking a question, then AI is understanding all these internal, the best amount of the documents and knowledge bases and databases and they can give comprehensive correct answers to the users. ... Then AI can actually give better answers and they can also give you another suggestion on what to do next. So, that is the biggest application.”
Additional applications include comparing labor agreements in the labor department or on the human resources side, screening several hundred resumes to find who is the best candidate for an open job requisition. Bonus point: it does all of this without the implicit bias that has been historically problematic for HR teams in their recruiting process.
Allganize is also able to create applications without traditional coding like Java and Python.
“Yeah, that is possible because of the LLM,” says Lee. “Because the LLM is such an amazing AI. So, it can handle most of the logics and data. It has been handled only by the software code. But now, the LLM doesn't cover these things. And so users, like somebody who wants to build any of those complicated workflow automations, they don't really need to write a code and they can just use LLM. And then they can just do visual flowcharts, like the old way. And then that is our app. And then all these things are done just like magic.”
Certainly, there’s been quite a bit of “magic” coming from Allganize and that’s all because of Lee and his innate ability to identify problems that need to be solved before they develop and create applications and solutions to address them ahead of time.
“I think two main things are necessary to get to this point with Allganize,” says Lee. “The first thing is expertise in AI, and the second thing is real domain knowledge.
“And I have been working in enterprise, including telecommunications and gaming a little bit more than 20 years. So, how enterprise is working and how the workflow is being defined and how the company and the teams are collaborating, and I have been with it for 20 years. In my academical background, it's all about AI. And I'm writing a code and seeing the very technical details of all these AI models.”
Lee learned a lot during and after selling his first company that he founded, which put him on the correct path to get the funding needed to realize his dream for Allganize.
“I really want to make this a company which can go long and then big,” says Lee. “And that's why we are working on going public. We are doing business in the U.S., Japan, and South Korea. And at the moment, Japan is our biggest market. So, we are going to go public in Japan in 2025. And from that point, we're going to accelerate our global expansion.”
