Automatically monitor regulatory updates to map to your internal policies, procesures and controls. Learn More

1558 Enforcement Actions in the U.S. over past 30 days


FTC enforcements decreased 55% over the past 30 days


SEC issued enforcements: $37,812,859 over the past 30 days


50 Final Rules go into effect in the next 7 days


49 Mortgage Lending docs published in the last 7 days


1670 docs with extracted obligations from the last 7 days


new Proposed and Final Rules were published in the past 7 days


11906 new docs in within the last 7 days


Considering RCM Solutions?  Here’s an RFP to get started.

EITL blog

Getting the most out of data used to mean relying on getting the data. Now, with Expert-in-the-Loop, human experts help you get the best data for your business.

Artificial intelligence continues to expand as a powerful presence in business, and in the compliance function, in particular.

That’s for good reason. 

A recent IDC study  shows that not using artificial intelligence for data gathering purposes results in a 10% loss in productivity for firms in the compliance and legal industry.

The same study shows that companies who are using AI today for data gathering purposes aren’t getting the most from their investment, however. That, IDC says, is primarily due to several highly solvable problems. 

This from IDC:

  • Around 28% of the AI/ML initiatives have failed. Lack of staff with necessary expertise, lack of production-ready data, and lack of integrated development environment are reported as primary reasons for failure.
  • Fairness, explainability, robustness, data lineage, and transparency, including disclosures, are critical requirements that need to be addressed now.
  • Large enterprises still struggle to apply deep learning and other machine learning technologies successfully. Businesses will need to embrace Machine Learning Operations (MLOps) – the compound of machine learning, development, and operations – to realize AI/ML at scale.

Merging Human Expertise Into the Machine Learning Equation

Enter Expert-in-the-Loop (EITL) machine learning technology. 

EITL is derived from Human in the Loop (HITL) machine learning applications, which adds human input into the data gathering process. Statistics show that AI algorithms are just 80% accurate. Add an “expert” human element to the mix, and that number soars upwards 100% to actual intelligence.

In a mainstream machine learning data gathering scenario, the algorithm produces the data, based on a prior initial assessment of what exact data is needed. Once that data is initially produced, human experts can assess the quality of the data, add their own assessments, and the algorithm can use those comments to improve the data and help maximize the data gathering process.

In a commentary in Forbes , Kayvan Alikhani, CEO and co-founder of, describes EITL in more specific terms.

The Expert In The Loop model comes with the facility to measure an individual’s aptitude and even document his/her errors for future reference. While HITL uses inputs from many individuals without considering their relevance in the decision-making process, EITL recruits experts specializing in particular domains for assistance with associated supervisory tasks. HITL uses the ‘Law of Averages’ i.e. if many people’s inputs are used, the average result is correct. It is the way to go if one is looking for quantity over quality. EITL, on the other hand, makes for easier audits, intelligent algorithms, and more confident AI assessments.

EITL in the Regulatory & Compliance Sector

EITL, which Alikhani calls a “next-gen” AI model is already making major inroads into the financial services sector, particularly so in compliance, regulation and data security. There, machine learning analysts work hand-in-hand with compliance professionals to close information gaps, eliminate mistakes, and readjust data algorithm-driven classifications. 

That not only helps gather superior data that can save companies potentially millions of dollars in compliance costs, but also significantly reduces regulatory risks as companies move forward in more aggressive and risk-laden financial markets.

Companies who leverage EITL machine learning technology don’t have to rely on hiring internal experts to assess and adjust data algorithms for them. Instead, they can rely on machine learning developers to handle that task for them. 

Consider, for example, an investment firm concerned about tagging specific regulations to specific asset managers for more accurate compliance. 

Instead of absorbing the costs of bringing in experts, provides its own network of contributors, a group of regulatory and compliance experts, to review business documents and add the proper tags. Better yet, the regulatory change management tool also learns and internalizes any tag changes and will automatically tag similar investment firm content needed down the road.

A Step Up for Machine Learning

Expert-in-the-Loop represents a data algorithm upgrade for compliance teams looking to collect new information and, after assessing the data with help from experts, help machine learning models improve over time using new expert-driven data.

That not only helps companies eliminate error-prone and time-consuming compliance tasks, it gives companies access to regulatory industry experts who can help compliance teams establish oversight over machine-learning data gathering models.

That scenario combines the best elements of technology and humane expertise for regulatory-minded companies operating in the AI realm.

Tags: , , ,