Aporia has been acquired by Coralogix, instantly bringing AI security and reliability to thousands of enterprises | Read the announcement

2024 AI & ML Report

Evolution of Models & Solutions

The process of setting up production ML environments and ensuring effective monitoring and observability is incredibly challenging in itself, before even reaching the stage of tackling other challenges relating to deploying new models.

Download Survey
December 23

Highlights

Real Time
88%

Of ML practitioners agree that real-time observability is crucial for the success of ML models in production

There’s a clear consensus among most ML engineers – regardless of their industry or team size – that real-time observability is a crucial requirement for the success of ML models in production, because without it, they are oblivious to any issues that may occur.

Production Issues
93%

Of ML engineers encounter issues related to
production models
on a daily/weekly basis

The fact that 93% of the respondents encounter issues with their production models on a daily/weekly basis highlights just how important it is to monitor and identify issues quickly, because a high volume of production issues can have financial implications for the business.

Hallucinations
89%

Of respondents report that their Large Language Models (LLMs) exhibit signs of hallucinations

With the increasing integration of LLM-based generative AI in enterprises, a major challenge arises: hallucinations, where AI produces inaccurate or nonsensical responses. This issue underscores the urgent need for solutions to ensure the accuracy and reliability of AI-generated content, essential for preventing reputational and financial harm.

Production
89%
Monitoring
83%
1 month
96%
Learn more

Observability isn’t just a tool for monitoring ML production issues, it also helps companies understand the effectiveness and impact of ML models from a use - case perspective.

Challenges

The top challenges respondents face with

their current ML monitoring system

32%

Setup and maintenance

30%

Integration with existing tools

29%

Limitations in providing value quickly

Download The Full Report

Fill out the form and get access to our 2024 Report