Find The Unknown

Mitigate risk, save time and cover more ground with purpose built digital assistants.

Truth Is In Digital Conversations
Hundreds and thousands of unstructured documents.
You need to find narrative related to specific topics quickly and across thousands of documents and digital interactions. This is what we have designed CompliancyAI to do!
Key Benefits
- Cover more ground - email, social, published articles
- User defined learning
- Save Time

How Do We Help?

Workplace Narrative Analysis

Garner projects that 47 percent of employers plan to let workers work remotely full time moving forward. In addition, 82 percent of business leaders across multiple industries plan to allow employees to work remotely at least some of the time as they reopen closed workplaces.

Unfortunately, 25% of women will experience sexual harassment in the workplace. Reality is that assessing thousands of employee interactions across emails, text messages and corporate chat sessions and finding relevant utterances is a long and arduous process, even if you are using keyword search based tools.

Empathy, understanding, and acceptance will drive the cultural workplace expectations.

You have new norms and expectations developing across your human resources and learning how your employees are accepting or rejecting certain norms helps build meaningful relationships with your employees. You have a need to learn the emotional side of your employees.

Here CompliancyAI custom narrative definitions come in handy. You define the narrative with examples of well known utterances as a list of three to four short phrases. CompliancyAI takes these definitions and automatically builds a graph of millions of nodes similar in nature or commonly used across the text corpus. What you get is a list of signals pending verification, the first step of getting key emails for review based on chatter related to narrative of interest just got easier. You don’t need to think of all possible keyword variations; CompliancyAI does that for you based on language usage in your data and your direction of relevant examples.

However, it is not always about the negative and hidden aspects. You enable feedback harvesting tools across mailbox, intranet forms/bots and let your employees tell you what is important to them, in confidence and anonymous. Powered with these insights, you can now understand the tribes in your organization based on transparent and honest feedback. This creates an environment where employees feel valued and heard - after all happy employees lead to happy customers!

Corporate Due Dilligence

As M&A activity returns, buyers will have concerns about their historical assumptions to value a seller in this environment. Buyers and their boards will be lot more cautious and will need more compelling reasons to stick with the deal.

a. Internal Due Diligence Documents – These will be less about sentiments.
b. Internal HR Risk Assessments – Internal Communications
c. Engagement and Client Experience – External Client Communications

Assess health of innovation, brand, and employee from internal corporate communications and documentation related to key business initiatives and projects. Understand the cultural ethos of the company through contexts related to evolving cultural norms, for example, "going to a meeting or"

Use customer chatter (both external and internal) as a proxy to assess the health of the brand. You can gauge key issues and product gaps reported by the users both online and across the support emails and chat message history. You will have the contrast in narrative before, during, and after the pandemic and an insight into the business's ability to delight the customer.

Litigation Document Review

Looking for the needle in thousands of documents powered with keyword search can be a long and tedious process. What if all you do is specify the top categories as a two or three word phrase list and let talkAItive do the work of mapping all similar utterances to your phrases of interest. This cuts down your time to get through the documents and identifies contexts which you would not have thought of, whether it is related and similar to your originally specified context.

Applying CompliancyAI to ENRON email data from January 1, 1997 - Jan 1, 2000, we set the goal of finding chatter related to liability and debt. The identified context related to the two search terms are signals deserving further validation. Across the 177K data points we set up a custom classifier which looked for narratives related to “off books”, “transfer liability” and “debt transfer” as key topics identified by our legal expert.

How Does It Work?

Identify Goals And Data
Learning goals and data sources like emails, documents, text, social and online chatter.

Run Analysis

Your data imported, cleaned and processed for relevant contexts.

Run Analysis

Your data imported, cleaned and processed for relevant contexts.

Review Contexts

Our insights provide top contexts across related docs for validation or adjustments.
Remove irrelevant contexts to your case for next refresh adjustment.

Test Hypothesis

Use variety of categories to learn related contexts with emotional variation across millions of data points.

Test Hypothesis

Use variety of categories to learn related contexts with emotional variation across millions of data points.

What does a full analysis include?


  • CSV/Text/PDF email files
  • Twitter, Facebook, Instagram, Google Reviews
  • GDELT, Stocktwits feed
  • Published Articles (online)
  • Other digital documents

Also Includes

  • 3MM-5MM datapoints approx 400K emails.
  • Pre-Trained workplace harrasment and anti-competition models
  • Re-train custom models
  • Access to UI dashboard for quick refrence and lookup.
  • Access to raw CSV output of all historical data and related insights.
  • Access to all insights through API to integrate into your own analytics stack.