HydroX AI is a company that is focused on providing comprehensive security measures, ensuring that your AI projects thrive in a secure and dynamic environment. Pulse 2.0 interviewed HydroX AI founder and CEO Zhuo Li to learn more about the company.
Zhuo Li’s Background
What is Li’s background? Li said:
“Prior to founding HydroX AI, I served as the Head of Privacy and Data Protection Office at ByteDance from December 2020 to February 2023. Before that, I held lead engineering roles at Facebook, LinkedIn, and HP.”
Formation Of HydroX AI
How did the idea for the company come together? Li shared:
“With the recent attacks that have happened through AI vulnerabilities, we feel it’s our responsibility to enable AI safety and build safe AI. By bulletproofing AI security, startups can reach their AI aspirations.”
Favorite Memory
What has been your favorite memory working for the company so far? Li reflected:
“I’ve worked with most of the people at HydroX AI before starting the company. It’s been cool working with smart, sharp people working on cutting-edge projects. These are problems that most people haven’t been addressing or aren’t sure how to tackle.”
“We wouldn’t be able to provide the solutions we’re providing if not for the experience everyone brings to the team.”
Core Products
What are the company’s core products and features? Li explained:
“As of right now, we have two more product features. We have a weapon store, of sorts. We have most attack methods in-house that we can leverage to attack Language Learning Models. We’re not aiming to jailbreak AI models to mess companies up but aiming to identify the issues. It’s elaborate and comprehensive, and it allows us to identify all risks.”
“This weapon store is proprietary to HydroX AI.”
“We have a pretty comprehensive dashboard to help people monitor and track the data. Some platforms already offer mitigation, but people also want to have data tracking and monitoring so they can know what happened in a visual and quantified way.”
Challenges Faced
What challenges has Li faced in building the company? Li acknowledged:
“We have been working with a ton of customers, but we’re working with one of the top leaders in the world. With big companies, it’s not uncommon that we need customization on our primary solution. We focus a lot of our bandwidth on these uniquely personalized projects, and the long term benefit would be that as we work with the giants in AI, we can replicate those success stories to customers of all sizes, working our way down to SMEs.”
“Big players know they need help with security, but they want to see that we’ll do what we say we’ll do. Our proprietary platform and arsenal allows us to show them their vulnerabilities.”
Evolution Of HydroX AI’s Technology
How has the company’s technology evolved since launching? Li noted:
“We built a very mature team and product in the very beginning, so there hasn’t been much since we began. However, in the AI space, a lot of new technology is coming out, with that, risks and vulnerabilities.”
“We’ve found a way to pipeline these innovations and their vulnerabilities, using our weapon store. With advancements made in AI, you begin to see a commonality in the vulnerabilities, so we abstract those vulnerabilities into fundamental and generic ideas.”
Significant Milestones
What have been some of the company’s most significant milestones? Li cited:
“1.) We know, very well, the safety and security of all the well-known models, and that allows us to give people a very good starting point of our demo.
2.) We launched our live evaluation platform, so before companies begin using a model they can consider what they see on our platform and decide whether or not this fits their needs.
3.) We’re capable of messing up any models in the world. That gives us the capability of identifying risks and vulnerabilities, so if we can tackle the best models, I would say we can tackle almost every single one in the world.”
Customer Success Stories
After asking Li about customer success stories, he highlighted:
“We work with a lot of SMEs, but we cannot mention some of the bigger companies we work with.”
Funding/Revenue
After asking Li’s funding and revenue information, he revealed:
“We raised $4 million in funding in our angel round of financing. This year, our goal is $6 million in revenue.”
Total Addressable Market
What total addressable market (TAM) size is the company pursuing? Li assessed:
“In 2023, the market amounted to around $200 billion and it is expected to grow beyond $1.8 trillion by 2030.”
Differentiation From The Competition
What differentiates the company from its competition? Li affirmed:
“We’re able to attack and identify the risks, making what we do real and not hypothetical. Mitigation is important, but we help companies safely detect their vulnerabilities without the detrimental impact it can have if their AI is hijacked in a real-world situation.”
Future Company Goals
What are some of the company’s future company goals? Li concluded:
“1.) We want to become the one-stop shop for risk control and data protection. Right now, we have products and components that we’d like to make easily accessible for anyone, democratizing the services we offer to bigger companies.
2.) We want to work with the top companies in AI and have them as our client success stories.”
