How Intuit uses LLMs to explain taxes to millions of taxpayers
AI makes tax language human again
Intuit's transformation of tax software through AI represents one of the most practical applications of large language models (LLMs) in everyday consumer technology. At a recent industry talk, Jaspreet Singh detailed how Intuit leverages AI to translate complex tax jargon into clear, personalized guidance for millions of TurboTax users—creating what they call "explanations as a service" that bridges the gap between IRS requirements and human understanding.
Key Points
- Intuit built a specialized LLM system that translates complex tax terminology into conversational, personalized explanations tailored to each user's unique situation
- Their approach combines multiple models working together: a routing model that understands user context, specialized explanation models for different tax domains, and quality control systems that ensure accuracy
- The system successfully handles the tension between providing simple explanations while maintaining complete technical accuracy—a critical balance in the highly regulated tax field
When AI speaks human, compliance follows
The most impressive aspect of Intuit's implementation isn't just the technical achievement of creating AI tax assistants—it's their recognition that technical accuracy alone isn't sufficient. People don't just need tax software that calculates correctly; they need explanations they can trust and understand.
This insight addresses a fundamental problem in financial services: regulatory requirements often force companies to communicate in dense, technical language that satisfies legal standards but fails to actually inform consumers. By developing AI that can maintain legal compliance while delivering truly understandable explanations, Intuit has found a way to enhance both customer experience and effective compliance.
The implications extend far beyond taxes. Banking, insurance, investing, and healthcare all suffer from similar communication challenges—where regulatory complexity creates barriers to consumer understanding. As LLMs evolve, we'll likely see this pattern of "translation as a service" become standard across regulated industries.
Beyond the transcript: AI's communication potential
What makes Intuit's approach particularly valuable is how it respects user intelligence while removing unnecessary complexity. Their system doesn't oversimplify tax concepts—it translates them appropriately based on the user's context. This personalized complexity management represents a significant advance over traditional "one-size-fits-all" financial explanations.
Consider healthcare as another domain ripe for this approach. Medical consent forms and insurance explanations of benefits are not
Recent Videos
Hermes Agent Master Class
https://www.youtube.com/watch?v=R3YOGfTBcQg Welcome to the Hermes Agent Master Class — an 11-episode series taking you from zero to fully leveraging every feature of Nous Research's open-source agent. In this first episode, we install Hermes from scratch on a brand new machine with no prior skills or memory, walk through full configuration with OpenRouter, tour the most important CLI and slash commands, and run our first real task: a competitor research report on a custom children's book AI business idea. Every future episode will build on this fresh install so you can see the compounding value of the agent in real time....
Apr 29, 2026Andrej Karpathy – Outsource your thinking, but you can’t outsource your understanding
https://www.youtube.com/watch?v=96jN2OCOfLs Here's what Andrej Karpathy just figured out that everyone else is still dancing around: we're not in an era of "better models." We're in a different era of computing altogether. And the difference between understanding that and not understanding it is the difference between being a vibe coder and being an agentic engineer. Last October, Karpathy had a realization. AI didn't stop being ChatGPT-adjacent. It fundamentally shifted. Agentic coherent workflows started to actually work. And he's spent the last three months living in side projects, VB coding, exploring what's actually possible. What he found is a framework that explains...
Mar 30, 2026Andrej Karpathy on the Decade of Agents, the Limits of RL, and Why Education Is His Next Mission
A summary of key takeaways from Andrej Karpathy's conversation with Dwarkesh Patel In a wide-ranging conversation with Dwarkesh Patel, Andrej Karpathy — former head of AI at Tesla, founding member of OpenAI, and creator of some of the most popular AI educational content on the internet — shared his views on where AI is headed, what's still broken, and why he's now pouring his energy into education. Here are the key takeaways. "It's the Decade of Agents, Not the Year of Agents" Karpathy's now-famous quote is a direct pushback on industry hype. Early agents like Claude Code and Codex are...