AI/ML — Disruptor and Enabler for EIA, Compliance oversight Market
This question hasn’t surfaced for the first time. Every time a new technology breakthrough takes place, there are FUDs(fear, uncertainty and doubts) about whether it would be a constructive or destructive change! This happened when digital adoption came glaring on the face of paperwork happy employees, When digital Supervision replaced manual one and is happening again when AI/ML models are looking potent enough to disrupt the traditional conditional logic based software for EIA and compliance oversight (monitoring, supervision, surveillance etc.)
Understanding the Product and the environment it operates in: The enterprise information archiving market is segregated into various verticals, including government and defence, Banking, Financial Services and Insurance (BFSI), education and research, healthcare and pharmaceutical, manufacturing, media and entertainment, retail and eCommerce, IT and telecommunications, and others (legal, electronics, automotive, construction, transportation and logistics, travel and hospitality, food and beverages, and energy and utilities). These verticals are expected to witness high adoption of enterprise information archiving solutions to achieve benefits, such as optimizing the storage information resources, lowering the risks, improving enterprise efficiency, and maintaining the transparency of the enterprise. Among these verticals, the BFSI vertical is expected to grow at the highest growth rate, due to the growing online financial transactions across the globe. Financial institutes and enterprises are continuously facing problems to meet regulatory compliance requirements defined by regulatory bodies. These regulations ensure that financial organizations work according to standards and maintain fair transactions with customers. In order to accomplish this, financial services organizations need to incorporate enterprise information archiving solutions, which can capture, archive, and provide easy access to the required information, enabling the organization to respond to agency information requests.
Asia Pacific (APAC) is expected to hold the largest market size in the global enterprise information archiving market. Many countries in APAC are increasingly moving toward the digitalization of their processes, resulting in a larger generation of data. There is a rise in competition in the enterprise information archiving market, due to the increasing enterprise information by large enterprises as well as Small and Medium Enterprises (SMEs). The enterprise information archiving solutions are playing an important role in overcoming the demand for storage devices and reducing the cost of the IT infrastructure.
Major growth factors for the market include reduced storage costs required for enterprise information archiving and government mandates to store enterprise information for audit and investigation purposes. On the other hand, the lack of awareness of the availability of enterprise information archiving solutions and heavy dependency on traditional approaches may restrain the market growth.
The major enterprise information archiving vendors operating in the market include Google (US), Microsoft (US), HPE (US), IBM (US), Dell (US), Veritas (US), Barracuda (US), Proofpoint (US), Smarsh (US), Mimecast (UK), ZL Technologies (US), Global Relay (Canada), Micro Focus (UK), OpenText (Canada), Solix (US), Archive360 (US), Everteam (France), Pagefreezer (Canada), Jatheon (Canada), and Unified Global Archiving (US).
Understanding the Users :
Like any other B2B product, the Buyers and Users of the compliance archiving solution are different. The buyers are chief compliance/technology/legal officers, whereas the users are the dedicated Reviewers, admins and People managers who are tasked with ensuring compliant communication in their organization. What we found out was that more than 50% of employees who are given the task of reviewing other employees are frustrated with how menial and laborious this job is. while they have many in-house and 3rd party tools at disposal, they have their own limitations, which leaves much to be desired. They have been hearing about cutting edge AI/ML-enabled tools and are very hopeful that it would ease their pain points.
Problem Statement: The problem statement of the survey was to find out whether the market is largely ready and comfortable with the idea of AI/ML over human intelligence for level-1 reviews. The Level-2 or escalated reviews, of course, have to continue with human reviewers for foreseen time.
Proposed Solution: The solution weighs on the below hypothesis: Hypothesis 1: If we introduce ML and NLP based scenarios, Our Supervision application will be able to reduce false positives as discovered by traditional policies.
Hypothesis 2: If we introduce ML and NLP based scenarios, they will be able to uncover more risks that are not discovered by traditional policies
The reimagined compliance monitoring product leverages AI/ML and aligns well with the defined problem. the inclusion of AI/ML and NLP models to create scenarios well suited to classify violations or non-violation reduces the effort of Reviewers to a great extent. After the passage of the beta phase, the reviewers would be convinced to not look at non-violations and rely mostly on the violations as flagged by the ML/NLP. This assisted focus helps Reviewers complete their reviewer quota and thus achieve business goals. With the scenarios integrated into the existing policy model, the new solution is well integrated into the current product and preserves intuitive interaction with the user.
The identified metrics like,
- %age of classification problems the AI/ML model is able to predict correctly
- effectiveness in identifying semantics and identifying semantics in non-English languages
- No. of different languages the model can detect and work upon
- Time taken by the AI/ML model vs the time taken by traditional policies
- No. of documents evaluated vs the traditional policies
, are used to measure to see how successful the solution is, are clearly defined, and a proper rationale is given.
The product road map, including both short-term and future goals, are explained as below:
- Q1: Open source pre-trained ML model to be customized for the Policies
- Q2: The customized model should be leveraged into multiple violation scenarios
- Q3: The scenarios should adhere to all MVP features and be feasible Nice to have features and dev testing to start
- Q4 : The system integration and end-to-end testing should qualify the model to be usable in customer UAT and PROD environments
