
Episode #33
How AI Gets Reliable Financial Data: Inside the AI Supply Chain
How does AI get reliable information—and how can businesses use AI to automate work, build products faster, and even create new revenue from their data? On this episode of Real AI Use Cases Business Owners Roundtable hosted by Elizabeth and Richard Gearhart , Andrew Lebbos, SVP of Licensing at Benzinga explains how a financial data company is using AI throughout its business while also supplying trusted financial data to AI platforms. Benzinga provides data to platforms including Perplexity, helping AI systems answer financial questions with structured, verified information. Andrew explains why data accuracy and quality assurance are critical for AI , especially in financial services where incorrect information can have serious consequences. Benzinga combines automated quality checks with human review to identify errors, outliers and AI hallucinations before information is published or delivered to clients. The conversation also explores how businesses are building around AI instead of simply adding AI to legacy systems , how AI is helping engineers dramatically reduce time to MVP, how employees across different departments are using AI differently, and why businesses should use enterprise AI tools rather than uploading company information into consumer accounts. Andrew also shares his experience using Claude, Perplexity, Perplexity Computer and Genspark , including an impressive example of using AI to create a presentation that would have taken approximately 40 hours to produce manually. Perhaps most importantly, Andrew explains how businesses with valuable data can potentially license that data to AI companies and LLMs —and why consistency, structured delivery, accuracy and credibility are essential to becoming a valuable data provider. TL;DL — Too Long; Didn't Listen AI isn't just a tool businesses use—it is becoming an ecosystem businesses can power. Benzinga uses AI internally, helps clients build AI systems, and licenses financial data to AI platforms. This episode explores AI automation, reliable data for LLMs, AI hallucinations, enterprise AI, AI productivity, data licensing and how businesses can prepare their data for the AI economy. What You'll Learn How Benzinga uses AI across its business and development processes How financial data can be used to power AI platforms and answer engines Why accurate, structured data is critical for LLMs How businesses can reduce AI hallucinations through data quality and human review How AI can automate tedious business processes Why some companies are building new systems around AI instead of adding AI to legacy technology How AI is helping engineering teams build products and reach MVP faster How employees in different departments can use AI in different ways Why businesses should use enterprise AI agreements for sensitive company information How businesses can potentially license valuable data to AI companies and LLMs What makes a data set valuable enough for an AI company to purchase How Claude, Perplexity and Genspark can be used for business productivity FAQs How is AI being used in financial services? AI is being used to automate processes, accelerate product development, analyze information and help deliver financial data through AI-powered platforms. Benzinga also supplies financial data to AI platforms. Why is accurate data important for AI and LLMs? AI models can make mistakes or produce hallucinations. For financial information, where there is often a specific correct answer, high-quality and verified data i... Chapters (00:00:00) - Introduction: Benzinga and AI Data Contribution (00:01:38) - Andrew's Role at Benzinga (00:02:28) - Building Systems Around AI (00:03:14) - Automating Podcast Workflow with AI Tools (00:03:57) - Ensuring Data Accuracy and QA Processes (00:05:10) - Government Trade Filings and OCR QA (00:06:30) - Favorite LLMs and Tools Used (00:07:42) - Company AI Usage Policies and Guardrails (00:09:02) - Team-Wide AI Usage Across Departments (00:10:07) - Perplexity Partnership and Data Licensing (00:12:06) - Business Model for Licensing Data to LLMs (00:13:58) - Reddit Data, Accuracy, and Financial Reliability (00:16:22) - Comparing Fable 5, Perplexity, and Other Tools (00:17:50) - Closing Thoughts and Benzinga's Mission






