# AI Can't Make Up a Number It Can't Look Up - Vinoo Ganesh, Kepler

## Executive summary

The core challenge in applying AI, particularly in finance, is not merely hallucination, but a structural lack of trust due to the inability to verify numbers and facts. The speaker details how traditional RAG platforms are insufficient because they fail to address the three sources of wrong numbers: retrieving the wrong document, extracting the wrong number from a correct document, or mathematical errors by the model. Kepler's solution is an architecturally different system that uses pre-processing to assign unique identifiers (pointers) to every fact, allowing the model to plan and point to these IDs rather than generating numbers itself. Verification is achieved through mandatory pointer dereferencing and post-processing checks, ensuring that every output fact is traceable to its source.

## Key takeaways

- The Core Problem: Structural Trust: The fundamental issue with current AI systems is structural. Simply stating that models will 'get better' to eliminate hallucination is nonsensical. For high-stakes applications, the system must be designed so that the output can be verified, corrected, or refused.
- Three Sources of Error: Wrong numbers originate from three sources: 1) Retrieving the wrong document; 2) Pulling the wrong number from a correct document; and 3) The model performing mathematical calculations (models predict tokens, they don't calculate).
- Kepler's Deterministic Architecture: The system operates by assigning pointers to unique identifiers to every fact during a pre-processing step (unstructured to structured data store). The model acts as a 'smart analyst' that plans the solution and retrieves data via these pointers, preventing the structural generation of numbers.
- Advanced Search and Security: To enable wide-then-narrow search, the system uses a combination of ANN + BM25 and cross-encoders. Security is paramount, requiring mechanisms to scope searches using OAuth tokens and prevent cross-tenant data leakage.

## Technical details

- Fact Extraction and Structuring: The system performs a giant pre-processing and contextualization step to convert unstructured data into a structured data store, assigning pointers to unique identifiers for every fact (e.g., a number in a 10K).
- Model Interaction and Verification: The model is restricted from structurally generating numbers; it must use a mechanism of 'dereferencing' the assigned pointer to output a fact. This is enforced during a post-processing verification step.
- Search Optimization: The architecture allows for wide search (using ANN + BM25 to get a set of candidate documents) followed by narrowing (using reciprocal ranking and cross-encoders) to overcome the inherent limitations of LLMs in broad searches.
- Security Scoping: Critical security considerations include filtering the tenant, preventing caller impersonation, and ensuring searches are scoped using mechanisms that constrain the set of documents returned with each query.

## Practical implications

- When evaluating AI vendors, builders must ask if the model writes pointers and if the output mechanism allows for verifiable output (i.e., clicking a number to see its source).
- The system must implement a mechanism for handling facts that 'come from nowhere' (hallucinations) by either correcting them at search time or refusing to output them.
- Building AI infrastructure requires moving beyond simple RAG and focusing on deterministic, verifiable scaffolding that enables trust.

## Topics

Generative AI, Financial Technology (FinTech), Retrieval-Augmented Generation (RAG), Data Verification, Deterministic Computing, Information Retrieval, AIE CODE summit

Source: https://www.youtube.com/watch?v=RdQCY_-SXbA
