By: Joshua Finley

Government data and regulations contain a wealth of information that can empower businesses and people to make more informed decisions. However, lengthy documents, complex legal terminology, and hours of video and audio make it extremely time-consuming to extract the most vital details. This is where David Martin Riveros, Managing Director of Iceberg Data, sees intelligent automation playing a pivotal role.

Autonomous Agents for Data Extraction

“There’s something called agentic workflows. This is when you give an AI web scraping tools or other capabilities, not just to answer questions, but to perform actions and run code on your behalf,” explains David.

He provides the example of an autonomous software agent that can analyze data by crawling government websites. The agent tries different strategies until it figures out how to extract the needed information, even if the site structure differs across agencies.

David contrasts this to static scrapers with predefined instructions. “We can give tools to an AI agent to use randomly until it finds a strategy to extract the data. So, they make decisions on how to gather and process that.”

Understanding Complex Data with AI

Machine learning has progressed rapidly in recent years. David reveals Iceberg Data is investing heavily in connecting their scrapers to AI solutions. This allows them to process and summarize lengthy government documents quickly.

“We do natural language processing to extract the context, tone, and intention behind the sentences in videos and text,” he explains. “The AI does an interpretation based on deciphering complex legal terminology and jargon used by officials.”

Their automated agents can scan transcripts, audio, video, and legal files. The system then provides briefs highlighting the key points a user may want to know.

“It saves thousands of hours of analyzing recordings and gives rapid, real-time updates,” says David. “So, stakeholders in different industries can integrate this technology into their processes to stay informed.”

Compliance and International Expansion

Iceberg Data goes beyond a generic web scraper, focusing on regulated sectors like healthcare and finance. David reveals they partner with law firms to ensure compliance when serving banks, hospitals, and similar customers.

“We are customizing these legal analytics tools to specific jurisdictions depending on the country,” he states. “This offers transparency and reduces information asymmetry for smaller companies that currently lack access.”

Geographically, Iceberg Data plans to expand in Latin America after soft launching in Mexico. They will collaborate with NGOs to adapt their government data AI to each nation’s legal system.

User-Friendly Dashboards and Predictive Models

In addition to back-end APIs, David shares plans to develop user-friendly dashboards and mobile apps. “We will also embed predictive models to forecast regulatory changes and government actions,” he says.

Their AI can customize reports visually and answer natural language queries from non-technical users. “There’s a way to ask the AI to present forecasts more appealingly. We’re researching how to have it not just provide text but analytical products as well,” explains David.

Ensuring Data Accuracy with Rigorous QA

With data-driven decision making, quality beats quantity. David reveals a robust QA process sets Iceberg Data apart from other scrapers.

“You have to sample the dataset, check values, and run tests to catch duplicates, missing fields, outliers etc,” he advises. “We even have our team act as end customers, analyzing the data before delivery.”

This validation ensures the scraped information closely matches an API directly from the source. David says customers appreciate their rigorous methodology after bad experiences with unreliable data.

“Other companies train their software with any data they find, often without proper quality assurance. As a result, they provide inaccurate insights to customers,” he warns. “We follow an enterprise-grade process to ensure the accuracy of our data.”

New Business Models Powered by Reliable Data

In closing, David reiterates the enormous potential of agentic workflows. “These AI tools allow people to take actions and decisions more transparently and efficiently regarding complex government data,” he remarks.

He invites stakeholders across industries to consider integrating this technology into their processes. “They could use AI to drive better, more accurate and timely decisions, so they’re aware as soon as something changes.”

David concludes, “If you can provide a trusted data source, you unlock the potential for others to build businesses on top of that. It’s the glue starting to bring processes across industries together.”

To learn more about David Martin Riveros and his approach, check out his LinkedIn profile, his personal website, and Iceberg Data’s website.

Published by: Nelly Chavez