Today, Website Planet‘s Roberto Popolizio had the chance to speak to
Liran Hason, CEO and Co-Founder of Aporia, an ML Observability platform that raised a $25 million Series A funding in February 2022.
They looked into the story behind Aporia’s early days, its evolution over time, and what is in the future of the company and
the web world in general.
Please Present Aporia To Our Audience. What Services And Products Do You Offer?
Aporia is the leading ML Observability Platform, empowering organizations to trust their AI and showcase the value of their models. By supporting all ML use cases and providing unparalleled customization, leading tech companies from the US and around the world – such as Bosch, Lemonade, & Armis – leverage Aporia to monitor, explain, and maximize the value of their machine learning models in minutes.
With Aporia, you’re able to centralize and visualize all your models in one place, making it easy to get the most insights and value from your models in production. It offers several key features, including single-pane-of-glass visibility through customizable dashboards, real-time monitoring of models in production with live alerts, explainable AI capabilities for simulating “what if” scenarios and understanding the impact of different features, and root cause analysis for investigating and debugging issues.
Aporia has raised $30M to date and experienced exponential growth, highlighted by significant expansion in the US. Organizations across different industries can leverage Aporia to ensure that AI can be adopted with confidence. With the expansion, the Aporia team has nearly tripled in the past year and continues to grow.
Can You Briefly Explain What Ml Model Monitoring Is And Why It Is Important for Online Businesses?
Machine learning models are first trained on specific data sets, however, once released to production, models start making impactful predictions for businesses and customers based on dynamic real world data. With this in mind, models can experience drift, decay, and we can see performance degradation over time, due to a multitude of reasons. Additionally, machine learning models are black box by nature, as we don’t know exactly how predictions are made.
ML Model Monitoring continuously tracks model performance to improve production models and ensure they are driving their intended value. Many online businesses employ machine learning models for various use cases, such as recommendation systems (RecSys), dynamic pricing models, demand forecasting, Customer LTV, and more. These models are making critical decisions for organizations that affects customer satisfaction and loyalty, and of course, their revenue. Without a dedicated, reliable ML monitoring system in place tracking and explaining model predictions, online businesses face the potential of losing profits and customer faith due to issues in their ML pipeline.
Not getting enough clicks? Challenged with rapid customer churn? Aporia’s ML monitoring helps online businesses conquer their AI goals, boost sales, and streamline business processes to create more clicks, more loyal customers, and more profits.
What Makes You Stand Out From The Competition?
Leading the ML Observability market, Aporia’s plug-and-play platform enables data science, ML teams, and business stakeholders to centralize and visualize all of their models running in production. Aporia’s suite of customizable observability tools ensures organizations maximize the value of their AI – with customizable dashboards that offer clear visibility into the revenue-impacting performance of the model, live monitoring and alerts, a way to explain model predictions with confidence, and a root cause investigation toolkit. With Aporia, online businesses are able to accelerate AI adoption and easily translate their predictions into revenue from day one.
Online businesses rely on quick decision-making and confidence in their model predictions. This is especially true for the robust eCommerce arena, where split-second purchase decisions based on highly curated recommendations generate a lot of their revenue. Aporia ensures your recommender system is driving clicks and helping your model make a bigger impact on your business.
What Are The Best Ways Marketers And Can Leverage Aporia?
Many marketers rely on Customer LTV to determine the dollar value of their existing and future customers or leverage Lead Scoring models to determine how likely a customer is to engage, purchase, or use their product or service. By using Aporia, marketers can confidently use their different models for their unique use cases to drive their marketing goals. There is an inevitability with machine learning models, at some point they will decay, the data will drift, or performance will degrade. Therefore, having a system like Aporia to continuously monitor and explain performance helps verify model predictions, amd ensure your model is generating revenues and impacting the business.
What Trends Are Going To Shape The Future Of Your Industry?
We’ve already started to see generative AI make groundbreaking changes to art and design. Now, we’re just touching the tip of the iceberg with the potential that models like ChatGPT offer. In the near future, generative AI will likely have a significant impact on a number of different industries:
- Healthcare: Generative AI could be used to design and discover new drugs, predict patient outcomes, and optimize treatment plans.
- Manufacturing: Generative AI could be used to design and optimize manufacturing processes, as well as to generate new products and materials.
- Education: Generative AI could be used to create personalized learning experiences, adapt to individual student needs, and generate new educational content.
- Art and media: Generative AI could be used to create new music, art, and other forms of media, as well as to analyze and understand existing works.
- Customer service: Generative AI could be used to generate personalized responses to customer inquiries and to improve the efficiency of customer service operations.
These are just a few examples of industries where generative AI is changing the game. This is the beginning of a new era – just like when Google changed online search. ChatGPT is revolutionizing the way we consume data and interact with the internet. It’s already shifting our mindset, but we are still far from realizing its full potential. Take a page from the movie “
Inception” and remember to keep a spinning top when diving down the generative AI rabbit hole.
Keep an eye out for the emergence of the ML product. Many times, there exists a gap between data science objectives and business KPIs. Embracing the ML product approach represents a fresh perspective on constructing ML models, shifting away from treating them merely as isolated models. Instead, consider them as comprehensive products designed with their utilization, objectives, and customer base in focus.
Lastly, Any Exciting News Or Developments That You Would Like To Share About Aporia?
We are always growing and looking for new people to join the Aporia team. So, if you’re a team player, want to learn more about the world of machine learning, and think outside the box – definitely check out our opportunities:
https://apply.workable.com/aporia/#jobs
