In the landscape of data analytics and predictive analytics, Hrvoje Smolic stands out as a pioneer with a vision. As the founder and CEO of Graphite Note, Smolic has dedicated the last two decades to pushing the boundaries of how we understand and leverage data. In a recent SafetyDetectives interview, he shares his journey from a business intelligence consultant to creating a startup that simplifies complex data science processes for users worldwide. Smolic discusses the inspiration behind Graphite Note, aiming to make machine learning and AI accessible to all through a no-code platform. His insights into the future of predictive analytics and the unique features of Graphite Note offer a glimpse into the next frontier of data-driven decision-making.
Can you introduce yourself and talk about what inspired you to enter the field of predictive analytics and data science?
Hi, my name is Hrvoje Smolic, and I’m the founder and CEO of Graphite Note. My journey in the data analytics space began shortly after my Masters in astrophysics; now it spans the last two decades, beginning as a business intelligence consultant, then transitioning into a data visualization specialist. In 2010, I launched my first startup, Qualia BusinessQ, focusing on developing a data analytics and visualization platform.
The field of data science caught my attention around 2015-16, when it started gaining significant traction globally. I was drawn to data science, seeing it as an extension of data analytics, but with a deeper focus on leveraging ML and AI algorithms across various use cases and industries. This realization came from my hands-on experience in several data science projects, which led me to ponder the possibility of creating a SaaS product that could simplify these complex processes for the end-user. This thought process eventually culminated in the birth of Graphite Note in 2020, a platform designed to democratize AI and predictive analytics through a no-code approach. We launched our MVP in 2021 and have since grown into a startup serving clients across four continents.
What sets Graphite Note apart from other predictive analytics platforms?
There are three key differentiators:
- Ease of Use: We aim to be the most user-friendly no-code ML/AI platform on the market. Our clients often tell us that with just “Excel user knowledge”, meaning a basic understanding of tabular data, they can effortlessly use Graphite Note to connect to their data and create various predictive models in a completely no-code environment. Some of our first clients didn’t even request a demo; they simply signed up for a 14-day trial and decided to continue with us thereafter.
- Comprehensive No-Code Machine Learning Models: Our platform covers 99% of current market needs with models for segmentation, regression, classification, timeseries and clustering. We also offer five specific models for e-commerce, retail, and SaaS companies that go beyond the usual scope, handling things like customer recency, frequency monetary, clustering for e-commerce and retail customers, Pareto analysis and customer lifetime value. These offerings set us apart from competitors, allowing us to cater to a wide range of use cases.
- Prescriptive Analytics: At Graphite Note, we offer something quite unique called prescriptive analytics, through a feature in our app named Actionable Insights. This is where the real value lies, as it not only analyzes data but also provides precise recommendations on actions to increase sales, improve features, or reduce costs. Our prescriptive analytics can, for example, advise focusing on specific demographics with a higher propensity to buy, based on comprehensive data analysis. This actionable guidance is what makes our platform truly stand out, offering direct paths to improve various aspects of a business.
In essence, Graphite Note is not just about predictive analytics; it’s about empowering businesses to make data-driven decisions quickly and efficiently, without the need for deep technical knowledge.
How do you see the role of no-code platforms evolving in the tech ecosystem?
No-code platforms are and will always be inherently limited because, by design, they don’t offer too much flexibility. If you introduce too much flexibility, you end up with not no-code but low-code or code platforms.
So, I think no-code platforms will continue to incorporate more use cases and provide greater flexibility than they do today for end users, especially the more advanced ones. However, there’s a limit to this because, as I mentioned, too much flexibility would transform them into low-code platforms, which defeats their initial purpose.
Thus, I foresee an evolution within no-code platforms, but it will be a balanced one.
In what ways has the approach to data analytics changed with the advent of AI and machine learning technology?
The approach to data analytics has changed in many ways. Essentially, data analytics revolves around crunching data, and having any form of assistance to make sense of this data is invaluable. Today, we have support from AI and machine learning technologies, which is help we didn’t have 10 or 20 years ago.
I remember, 20 to 25 years ago, I would spend a week writing Java code to create a batch report for bank statements. That was a significant part of my job. Then, data visualization tools like Tableau or Power BI came along, and suddenly, you didn’t need a week to create a report. You needed just a few hours to drag and drop elements on your screen.
This was a tremendous leap forward thanks to data visualization tools. Now, no-code AI tools are set to change the landscape once again. Instead of data scientists spending weeks or months creating models, cleaning data, and identifying the most accurate algorithms, you can use no-code AI tools to accelerate this process and have results in a few hours. This is a revolutionary shift that’s happening right now, thanks to AI and machine learning technologies.
How do you foresee the integration of predictive analytics with other emerging technologies such as blockchain or IoT?
I had higher expectations for the integration of data analytics with blockchain and IoT. I explored some possibilities 7-8 years ago, but this integration hasn’t progressed as quickly as I anticipated.
However, I’ve always believed that any kind of analytics, especially predictive analytics, would be incredibly beneficial for both blockchain and IoT. For blockchain, one challenge is how to quickly read the databases. For IoT, gathering data from various devices presents a vast opportunity for predictive analytics, especially in predicting failures or in the predictive maintenance area. Although the integration hasn’t advanced as fast as I hoped, I believe this decade will see a stronger connection between AI, predictive analytics, and technologies like blockchain and IoT.
What challenges do companies face in scaling their analytics capabilities, and how can they address them?
The biggest challenges companies face today have already been mentioned. Whether it’s Shopify data or any e-commerce data, ERP, CRM, or accounting data from platforms like QuickBooks, every company, no matter how small, has digitized to some extent because they usually operate on some marketplace or SaaS product. So, data availability is not the issue.
The challenge lies in making sense of this data. Hiring a data scientist is problematic due to their scarcity and the high costs involved. In the U.S. alone, the market lacks two to three hundred thousand data scientists, and their services are very expensive, which not every company can afford.
The solution for companies to scale their analytics capabilities lies in adopting a no-code approach. This allows them to test hypotheses and play around in a safe, no-code environment. They can connect to their data and create models within hours, not weeks or months.
So, for any company looking to start with AI, going the no-code route is advisable. It offers a way to quickly understand what you can achieve in a few hours, rather than waiting weeks or months and spending a fortune on data scientists to do the job for you.