“There are many challenges in adapting a consumer product like ChatGPT to the enterprise – such as data privacy, hallucinations, and consistent answers – so we help consult with a lot of clients in navigating this transition.”
Hi Ivan, Welcome to our Interview Series. Please tell us a little bit about your role and responsibilities in your current company.
I’m the CEO and Founder of Datasaur, an NLP data-labeling platform that helps annotators train AI algorithms efficiently. I work with my teams to identify challenges our customers face with NLPs and develop solutions to support this process.
What was the idea behind starting an NLP training company?
I worked with data annotation initiatives at my previous three companies – most recently at Apple. At each company, we were stuck using a combination of spreadsheets, in-house tooling, and open-source tools. It was frustrating and inefficient. Rather than solve this problem one organization at a time, I decided to start Datasaur to solve this issue for the broader NLP community.
Please tell us about your Ideal Customer Profile (ICP) and the industries you are offering your services to?
The ideal ICP is a data scientist or AI engineer building NLP solutions. We serve a large swath of industries – anything from legal to finance to healthcare to e-commerce. Language underlies everything we do, and there’s a lot of unstructured text and audio data lying around. Some of our customers today include Google, Netflix, Zoom, Qualtrics, Stanford, Harvard, Oxford, and more.
How has the role of an AI-enabled CEO evolved in the recent months?
What are the profound changes you witnessed in the NLP modeling landscape since the arrival of LLMs/ generative AI models.
We are seeing a renewed wave of interest in everything NLP. Everyone is trying solutions like ChatGPT at home, realizing how far this technology has come, and wondering how they can leverage these capabilities in their day-to-day work. There are many challenges in adapting a consumer product like ChatGPT to the enterprise – such as data privacy, hallucinations, and consistent answers – so we help consult with a lot of clients in navigating this transition.
Please tell us about your recent funding and how you plan to expand your products/ services?
We are using the new funding to help our clients further automate their labeling and integrate it with their existing tech stacks. Everyone is focused on reducing their budget and operating efficiently in the current economy, and Datasaur can save a significant amount of time and effort in the data labeling process. We are also expanding our product capabilities to support gathering the training data required to finetune and customize LLMs.
Could you tell us about the toughest decision that you had to make in your AI career?
One thing that surprised me is that there are many product decisions that AI product managers make in a day’s work that would usually require months or even years to legislate. For example, I had to determine what types of content were “family appropriate” for a product with hundreds of millions of daily users. This is, to be honest, too much power in the hands of an individual. I’m glad to see that there is a lot of discussion around AI safety and data privacy to bring much needed perspectives to the situation and help standardize what is deemed acceptable.
How big is your AI development and marketing team?
Please tell us how you plan to expand your team size in the coming months?
Our team as a whole just reached 55 people. We will continue to hire intentionally over the next 12 months with a focus on profitability. We will likely be growing our customer success and sales departments next. One thing that’s unique about our team is that our product, design and engineering teams are all located in Indonesia, allowing us to leverage some of the top untapped talent there.
What kind of talent are you specifically looking out for your organization in 2023-2024? What have been the challenges in talent management in your AI startup?
The hardest challenge we face, and I have heard repeated from many others, is balancing our current responsibilities while keeping up with all the rapid developments in NLP this year. It’s truly a thrilling time to be working on NLP in the midst of one of the largest technological shifts of our lives.
Lighter notes:
Burn the midnight candle or soak in the sun?
Burn the midnight candle all the way.
Coffee, or Tea?
Both, but I start every day with coffee!
Your favorite Datasaur people management initiative that you want everyone to know about?
We offer shares to every employee in Indonesia. This is not a common practice there, and I often have to sit down and explain what the shares mean and how they work. But it’s important to me to have everyone truly feel like this is their company.
First memorable experience in your career as an AI leader?
Replacing a decades-old search algorithm with an ML model in just 3 months. It really opened my eyes to what was possible with this technology.
One thing you remember about your employee(s):
Meeting 40 new employees for the first time when I got to travel back to Indonesia post-COVID! It was so fun seeing everyone when we had only ever chatted on Slack and Zoom.
Most useful app that you currently use:
McDonald’s app has some great deals 🙂
Thank you, Ivan! That was fun and we hope to see you back CIO Influence soon.
[To participate in our interview series, please write to us at sghosh@martechseries.com]
Ivan Lee graduated with a Computer Science B.S. from Stanford University. He was chosen for the selective Mayfield Fellows entrepreneurship program in 2010 to teach students how to lead principled high-growth tech ventures. Ivan went on to co-found Loki Studios, an iOS game studio. After raising institutional funding from DCM’s A-Fund and launching a profitable Pokemon-style game with 1 million users, Loki was acquired by Yahoo! Inc.
Datasaur leads the NLP industry with its comprehensive and automated data labeling solution. Founded in 2019 and headquartered in Silicon Valley, the company helps financial, legal, and healthcare companies turn raw unstructured data into valuable ML datasets.
