Introducing hila Conversational Finance's New Disambiguation Feature

February 10, 2025

Smarter Queries, Clearer Insights

Financial analysis is often complex, and even the simplest questions can carry multiple interpretations. When users interact with financial data using natural language, ambiguity is inevitable. What does "current" mean? Are we talking about today, this month, or this fiscal period? What exactly falls under "income" or "margin"?

hila Conversational Finance (hCF) has always made it easy for users to access financial insights with simple, intuitive queries. Now, with our new disambiguation feature, Conversational Finance is even smarter — helping you clarify your intent so you get exactly the data you need, without the guesswork.

What is disambiguation?

Disambiguation in hCF refers to the process of clarifying and refining a response to ensure it accurately addresses the user’s intent. In the context of AI and financial planning & analysis (FP&A), it means:

  • Identifying ambiguous or unclear questions and prompting the user for clarification.
  • Distinguishing between multiple possible interpretations of a query to provide the most relevant answer.
  • Refining responses based on context, domain knowledge, and additional inputs to avoid misunderstandings.

This ensures FP&A professionals receive precise and actionable insights rather than generic or misleading answers.

How disambiguation works

Instead of making assumptions that might not align with your intent, hCF now detects when a question could have multiple meanings and requests clarification before running a query. This ensures that users always receive the most relevant and accurate responses.

Here’s how it works in action:

Example 1: Clarifying terms

User asks: What’s my cost with acme in 2024?  

Acme could mean a supplier, a customer or otherwise. hCF correctly interprets the lower-cased “acme” as a customer and finds the total cost.  

A screenshot of a computerAI-generated content may be incorrect.

Example 2: Explaining how a question was interpreted

User asks:Compare top 10 customers sales fy22 vs fy23

What does an FY refer to? While a financial user, or anyone in the business world, would naturally understand it as “fiscal year,” this is not a trivial translation for a large language model, especially one that then can use that query to generate SQL.

A screenshot of a computerAI-generated content may be incorrect.

Example 3: Interpreting shorthand

User asks:What’s the sales for Amber in 2023?

Does "Amber" refer to a product, a customer, or something else? hCF provides a helpful assumption and asks:

The query assumes that "Amber" refers to a customer whose name contains "Amber" (customer_name ILIKE '%amber%'). Is that correct, or are you referring to a product?

This helps avoid misinterpretations and ensures that users retrieve data relevant to their needs.

A screenshot of a computerAI-generated content may be incorrect.

New users, ready-to-go analysis

In addition to the disambiguation feature, hCF provides pre-built financial analyses for new users to get started quickly. This feature offers insightful reports right out of the box, showcasing the tool's capabilities and helping users understand what’s possible. By exploring these ready-made analyses, new users can:

  • Quickly familiarize themselves with financial data queries.
  • Identify key trends without needing to configure reports from scratch.
  • Gain immediate value from hila's insights without a steep learning curve.

The Impact: More confidence, better decisions

With the disambiguation feature and ready-to-go analyses, hila Conversational Finance doesn’t just retrieve data—it collaborates with you to ensure that every insight is precise and useful. By reducing ambiguity and providing instant access to key financial reports, finance teams can:

  • Save time by avoiding incorrect assumptions and re-running queries.
  • Gain confidence in the accuracy of financial reports.
  • Improve efficiency by making data retrieval smoother and more intuitive.

Smarter queries start today

Disambiguation and pre-built analyses are now live, ensuring that your financial questions lead to the right answers—every time. Try it out and experience a new level of clarity in financial decision-making.

Got questions? Get in touch.