Ever wished there was an easier way to bring your ideas to life? Imagine a tool that seamlessly translates markdown-style code into detailed diagrams, streamlining complex diagrams and flowcharts.
That’s precisely where Mermaid Chart comes in! The tool is redefining diagramming with AI precision and real-time collaboration.
Welcome to an exclusive edition of our Interview series on AllAboutAI.com, where we introduce various stakeholders and developers in the AI industry to share their stories, challenges, and insights.
Today, we enjoyed speaking with Dominic DeWalt, Business Strategy and Operations lead at Mermaid Chart. This groundbreaking AI-powered tool easily creates complex diagrams from markdown-style code and collaborates with team members in real-time.
This conversation is all about Dominic’s journey, the development of the Mermaid Chart, and the future of AI in Diagraming.
Be sure to watch the full interview on our YouTube channel.
Now, let’s get into the fascinating insights Dominic shared with us about his journey, the inception of the Mermaid Chart, and the transformative potential of AI in diagramming.
Please briefly describe your background and how you entered the tech industry.
I ended up here not by chance but through some random occurrences. I graduated from college in 2021 and started working at a real estate tech startup straight out of college. Unfortunately, that startup ran out of money, but I handled operations, account management, marketing, and generalist tasks during my time there, which inspired me.
After that, I joined a property management software company working on business systems. I acted as a bridge between the company and the engineers. Whenever something needed to be done, like automations or other tasks, I would help define what needed to be done and how best to do it.
I wrote all the requirements and then worked with the engineers to implement them. This was my first step into understanding the product management process, though I wasn’t exactly working on the product itself.
What inspired you to be part of Mermaid Chart?
I was there for about a year before moving to Mermaid Chart, a different landscape. We do diagrams with text, which works very well with AI compared to Appfolio, where I previously worked with property management software and just started with AI. While I didn’t have much experience with AI at Appfolio, the past six or seven months at Mermaid Chart have been a roller coaster.
AI is revolutionizing every industry, and there is so much to be done with diagramming. You can have AI do anything. The text-based style we use to create diagrams works very well with AI, so although I didn’t initially have an interest in diagramming or making charts, I shifted from the other industry to this one.
What were some of the biggest challenges you faced while developing Mermaid Chart, and how did you overcome them?
The biggest challenge is ensuring that the AI output is accurate.
The biggest challenge so far is that everyone wants to skip as much work as possible. When making a flowchart, they often dislike the entire process. AI is close to being very good, but the biggest challenge we face right now is ensuring the output is accurate. For example, you can paste an entire YouTube transcript into the AI and ask it to create a flowchart with Mermaid syntax. It’s pretty good, but there are still some inaccuracies.
We’re still working on the trust factor because Mermaid Chart, as a platform, is somewhat complex for non-technical people since it is text-based and uses Markdown-style text that looks like code. People unfamiliar with it are often intimidated when they first encounter it.
Therefore, we need to ensure that the AI output is as accurate as possible so that people can trust it and use it with minimal tweaks. If someone gets a significantly off diagram, it doesn’t help them much. We are getting a lot closer to achieving this accuracy and figuring out how to make it easier for users to make the output 100% accurate.
Is Mermaid for everyone… can students use it for assignments?
Yes. Mermaid Chart is designed for a broad audience, including non-technical users. However, the text-based aspect can be intimidating, so we’re working on making it more accessible.
Can you explain Mermaid Chart and how it differentiates from other diagramming tools?
Mermaid Chart is a tool that creates diagrams from text, which works very well with AI. Unlike traditional diagramming tools, the text-based style we use to create diagrams sets us apart because it integrates seamlessly with AI capabilities.
Mermaid’s text-based system is better suited for generating outputs from AI. While tools like Lucidchart or Miro have some AI capabilities, their outputs often don’t match what you asked for. AI models like OpenAI and Claude can write Mermaid syntax much more quickly than they can produce images or flowcharts directly. Translating the given context into Markdown-style code makes the process faster, easier, and more accurate.
Mermaid Chart stands out because of its text-based approach, where the diagram is entirely tied to the text. Unlike drag-and-drop whiteboard tools, this rigidity actually improves the accuracy of AI-generated outputs. As a result, our diagrams are much more precise than those from other whiteboard-style tools.
What challenges does Mermaid face as a brand and tool in the industry?
Mermaid is super interesting because it started as an open-source project about 8 years ago. The founder of Mermaid Chart needed a better way to create diagrams while using Visio, and with the emergence of Markdown as a new style, he saw an opportunity.
He started the open-source library, which took off because the Markdown style works so well. You can store it in Git and use it in GitHub, Notion, or Markdown note apps like Obsidian. The diagrams essentially design themselves.
What’s unique is that you don’t have to create visuals manually; you can quickly write a simple diagram. About a year ago, the founder received investment from Open Core Ventures, Sid Sabrangi’s venture fund. Sid is the CEO of GitLab and invests in popular projects to help them grow. This investment led to the creation of the Mermaid Chart.
Now, we are integrating Artificial Intelligence I into the Mermaid syntax and the Mermaid project. We’ve developed a better and new visual editor, so people can make changes by clicking instead of just using text.
However, a significant challenge is differentiating Mermaid Chart from the open-source software, which is very prominent. Mermaid Chart is new, so we must work harder to establish its identity.
Another challenge is expanding to a broader audience. Non-technical people are excited about self-designing diagrams using AI, but the Markdown-style text can be intimidating. Software engineers can pick it up quickly and run with it, but non-technical users often find themselves staring at a wall of text with no idea what to do. They can learn it, and it’s pretty easy to pick up, but it’s not intuitive.
We are trying to cater to various use cases. For example, some people visualize code bases, while students use it to visualize their lectures by posting their notes. There are many use cases for Mermaid Chart and AI, and we are figuring out which are most impactful so we can focus on those areas. The broad range of uses is both an opportunity and a challenge.
How does Mermaid Chart enhance team workflows and documentation processes?
The Mermaid Chart platform is excellent for organizations and teams because it allows real-time collaboration. You can create documentation, and team members can add comments, make notes on your diagram, and even build it with you. Everyone can do the same editing, which makes it very collaborative.
For documentation, one user might tackle the project independently, doing the first pass and highlighting areas where they need help. Others can then come in, see all the comments and notes, and understand the structure and syntax of the diagram. This provides excellent insight into how someone works, which is particularly useful for managers. It’s also beneficial because you can check all the work and make changes asynchronously without being on a call.
One significant advantage is that you can build and communicate asynchronously without ever leaving Mermaid Chart. It’s interactive and allows for seamless collaboration.
What is your approach to ensuring Mermaid Chart remains secure and reliable for enterprise use?
Security is a huge priority for us. We encrypt all diagrams, so neither we nor anyone else can see your content. Everything is encrypted on our end to ensure privacy. We offer on-premise installations for users with high-security needs, such as government contractors, which means Mermaid Chart can be installed directly on your servers. This ensures that no one else can interact with your data.
Additionally, we are obtaining our SOC 2 attestation to demonstrate our commitment to security further. Customization is also possible, allowing Mermaid Chart to be integrated into various systems to meet specific security and privacy requirements.
Could you share any notable success stories or case studies from organizations implementing the Mermaid Chart?
One of the most exciting conversations I’ve had with a user involved helping him use AI to document his entire codebase. He had 5,500 lines of code and didn’t want to read through it all to figure out what everything was doing. He wanted a quick solution to hand over to new engineers. He could paste the entire codebase into Plot AI, generating a massive flowchart. This saved him countless hours of work.
Another example, albeit a more personal one, was when I used Mermaid Chart to find anniversary gifts for my girlfriend. I input the context, such as living in Santa Barbara and her interests, and asked the AI to create a flowchart of gift ideas and where to get them. It provided specific stores and suggestions, saving me time and effort. This shows the versatility and practicality of our diagramming tool in unexpected ways.
What trends in the tech industry are you most excited about, and how do you think they will impact tools like Mermaid Chart?
I’m really excited about building our Large Language Model (LLM) and integrating various models like OpenAI, Claude, and Lama into Mermaid Chart’s AI. We’ve had users wanting to record their call meetings and instantly generate diagrams from them, so the trend of voice-to-AI is something I’m very interested in.
In the near term, we will integrate different models into Mermaid Chart’s AI, selecting the best one for each use case. We’ll also develop our own LLM to fill in gaps, ensuring outputs are more accurate, professional, and presentation-ready. We aim to help users streamline their processes with better AI and voice integration.
How responsive is the Mermaid Chart to user feedback? Can you share an example of feedback that led to a practical change?
I am the sole product owner at Mermaid Chart, so I love receiving feedback. We are very responsive and constantly seek user input. For example, we recently heard that our outputs looked too cartoonish. While it stung a bit, it was valuable feedback. Users want presentation-ready diagrams they can comfortably show to executives.
Traditionally, the Mermaid Chart has been used by engineers who need quick solutions, and it is often used in VS Code. We’re now moving towards creating professional, presentation-ready diagrams. We are making lines cleaner, nodes sharper, and changing colors to meet professional standards.
We’re currently bouncing these changes off users to ensure they meet expectations. While we’re a small team and can’t always address niche requests immediately, we aim to build for the people and tap into new markets continuously.
Five or three years down, where do you see the Mermaid Chart?
When I think about the Mermaid Chart, we want to tackle AI first. We aim to make the output as perfect as possible, but that’s not the long-term vision. While you can use diagrams for documentation, planning, and many other purposes, there are also reasons some people might avoid them, such as feeling that documentation slows them down.
We aim to help people create charts that facilitate documentation and transform those diagrams into something actionable. We want to execute processes and workflows, assisting users in building, planning, and implementing their processes simultaneously.
The long-term vision is to solve the tedious documentation problem and enhance efficiency in executing those plans. We want to streamline the entire process, from documentation to execution, making it more efficient and less burdened.
How do you measure success for Mermaid Chart, both in terms of business metrics and user satisfaction?
From a business perspective, we focus on metrics like how often people start a trial after signing up, conversion rates, and whether they enjoy the trial enough to sign up afterward. These metrics help us understand how we deliver and show value. Of course, revenue and other financial metrics are important too.
Measuring customer satisfaction is trickier since it often relies on user feedback. We try to gather insights from the usage database to see what features are working well and which aren’t. For example, if we release a new feature for our visual editor and see a drop in usage, we know it’s not resonating with users and can improve it.
We also track customer satisfaction through brief surveys during cancellations, asking users for feedback on how we can improve. Sometimes, users cancel because they don’t diagram often enough to justify a subscription, which makes sense. However, this feedback helps us understand if there are areas we need to fix, such as AI output accuracy. By addressing these concerns, we can improve the product.
Overall, the more feedback we receive, the better we can enhance user satisfaction. We continuously monitor user feedback and make improvements based on it.
If you could diagram anything in the world with the Mermaid Chart, what would it be and why?
Oh man, I feel like I already used my most exciting use case with the anniversary gifts. However, I think a dream scenario would be to plan the company’s vision and strategy using AI as a counterpart. Rather than just inputting everything I know, having the AI fill in the gaps in my knowledge would be amazing.
For example, I could come to the AI and say, “This is our strategy as a company. This is where we think we fit into the ecosystem.” Then, the AI could make suggestions and recommendations on what can be done. That would be my dream for Mermaid Chart—having a counterpart that helps us address our problems and achieve our goals, creating an entire plan that’s accurate and actionable.
If you could collaborate with any other tech company or startup, who would it be and why?
GitHub is our dream collaboration.
Well, I wouldn’t call them exactly partners or anything yet, but GitHub is our dream collaboration. There are many co-pilot opportunities with GitHub, and Mermaid syntax is already usable there. You can insert it quickly, but we aim to get into the GitHub co-pilot program.
GitHub recently announced they’re working on co-pilot extensions similar to ChatGPT’s GPTs. Being an extension of GitHub, integrating more diagrams and documentation into people’s work and doing it automatically would be a dream collaboration for us. We want to work with anyone and everyone if we can, but GitHub is definitely at the top of our list.
We’re working towards integrating Mermaid Chart with GitHub co-pilot to enhance documentation and diagramming capabilities within the GitHub ecosystem.
What’s the most unusual or creative use of the Mermaid Chart you’ve seen?
This is a really interesting one. I talked to one guy who wanted to use Mermaid Chart to map the New York City subway system. Our diagram format resembles a metro map with lines and stops along the way. He wanted to use this as a test to build maps for his Dungeons & Dragons (D&D) campaigns.
For his D&D campaigns, he wanted to map out where the campaign would start and the cities along the way and use that to plan the entire campaign. It’s probably the most unusual and creative use case I’ve heard from a user. Seeing how our tool could be applied in such a unique context was fascinating.
What advice would you give to aspiring entrepreneurs looking to create their own tech startup?
There’s no better way to figure out what to build: to get advice from people who have already done it and get feedback directly from the people you’re creating for. You don’t want to make something and then try to sell it to a group of people without ever talking to them.
It’s much easier to say, “This is what you want, this is what you need, these are your problems,” than solve something for them. Listening, always learning, and gathering feedback as much as possible is crucial.
Regarding AI, I would advise testing all the models available for your specific use case. Figure out which one works best for what you’re trying to achieve. Improve prompt engineering and determine which output aligns with your goals. Implement that model and work on improving your system even better.
How can our readers stay updated with the latest news and updates from the Mermaid Chart?
There are a few different ways. You can check out mermaidchart.com, and once you get into the system, we’ll send you regular product updates. We also have a Discord community if you’re interested in joining it.
For the latest updates, you can follow us on our new podcast read our articles, and blog posts. You can keep up to date by following us on LinkedIn as well. Of course, you can contact me directly anytime—I’m super happy to help.
Anything you would like to say to our readers and Potential clients?
Yes, if you’re interested in diagramming or AI in this space, please get in touch with me. I would love to hear all of your thoughts and feedback. You can reach me on LinkedIn or our website, and I will be there to help.
Key Takeaways
Our conversation with Dominic DeWalt offered deep insights into the innovative potential of Mermaid Chart, highlighting how this AI-powered tool is transforming the way we create and interact with diagrams. From its text-based approach that ensures precision to its capacity for real-time collaboration, Mermaid Chart is set to revolutionize diagramming across various industries.
As AI advances, tools like Mermaid Chart will undoubtedly play a crucial role in enhancing productivity and creativity. For the latest updates and to see Mermaid Chart in action, follow their journey on their website and social media platforms.
