Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Anne Griffin built a cyber-risk agent that called a data science model to estimate vendor risk from similar companies. Generative AI was too probabilistic for a use case where accuracy mattered, so the agent combined that model with LLM output rather than relying on the LLM alone.

Related Insights

While generative AI models can hallucinate with low stakes, industrial AI cannot afford errors. This has created a premium for companies with unique, real-world datasets that are verifiable and critical for high-stakes decisions where failure could be catastrophic, like an explosion.

An effective AI agent isn't a single, all-knowing model. It's custom software that uses LLMs for inference only when necessary, relying on cheaper, deterministic code for most tasks. The goal is to maximize outcomes, not token usage, by blending AI with traditional software.

EliseAI predicts lease renewals with an in-house neural network, not an LLM, because it has lots of behavioral data and needs reliable probabilities. Martone says LLMs struggle with determinism and probability, so his team picks the tool based on impact and shipping speed rather than defaulting to LLMs.

While content generation is impressive, the highest value for financial professionals lies in using AI as a verification layer. A tool that can audit a complex model and catch a single, costly mistake provides more immediate ROI than one that simply builds the model from scratch.

While generative AI captures headlines, organizations gain more immediate and reliable value from traditional machine learning (predictive AI). Its deterministic nature provides consistent, repeatable outputs from quality data, making it the foundational backbone for scientific AI applications.

For critical enterprise functions like financial modeling, 99.9% accuracy from a probabilistic LLM is unacceptable. Platforms like Salesforce's Agent Force 360 solve this by layering deterministic logic and guardrails on top of the AI, ensuring compliance and preventing costly errors where even a 0.1% failure rate is too high.

Building reliable AI agents for finance, where accuracy is critical, requires moving beyond pure LLMs. Xero uses a hybrid system combining LLM-driven workflows with programmatic code and deep domain knowledge to ensure control and reliability that LLMs inherently lack.

Pega's CTO advises using the powerful reasoning of LLMs to design processes and marketing offers. However, at runtime, switch to faster, cheaper, and more consistent predictive models. This avoids the unpredictability, cost, and risk of calling expensive LLMs for every live customer interaction.

To deploy LLMs in high-stakes environments like finance, combine them with deterministic checks. For example, use a traditional algorithm to calculate cash flow and only surface the LLM's answer if it falls within an acceptable range. This prevents hallucinations and ensures reliability.

To solve for LLM non-determinism, a hybrid approach first uses an LLM to evaluate new agent behaviors. It then analyzes these interactions to auto-generate specific, deterministic rules. Over time, this shifts most traffic to a fast, reliable rules engine, reserving the LLM only for true novelties.

Use data science models for high-stakes risk, not generative AI | RiffOn