Foodservice operators rarely have two days that look the same. New challenges constantly emerge, and making dozens of decisions with limited time and incomplete information is the norm: Why did one location's food cost suddenly spike? Which limited-time offer is truly driving profit? Why does one location consistently outperform another?

Finding these answers often means sifting through multiple reports, exporting spreadsheets, and piecing together data across different systems—a time-consuming process that pulls operators away from their core role of hospitality.

Data presentation and pattern recognition are where AI excels. However, as Matt Sundberg, Senior Vice President of Products at Craftable, points out, when restaurants use generic AI models, they lack the critical context needed to deeply understand their own operations. This is why Craftable launched its AI model, "Crafti." Craftable is a software platform that replaces fragmented spreadsheets and manual processes with automated, real-time visibility into food costs, supplier spend, and operational margins.

"Crafti's conversational layer allows general managers (GMs) to ask questions at any time and receive grounded answers with a path to resolution, without waiting on colleagues or submitting tickets," Sundberg said.

From Reports to Conversation

With conversational AI, GMs can interact with operational data in a natural way, uncovering insights that improve decision-making across every part of the business.

"Operators are surprised to find that Crafti doesn't just provide answers—it drafts solutions," Sundberg said. For example, if a GM asks to describe next week's order, Crafti drafts a purchase order based on actual supplier catalogs, pricing, and pack sizes. If given a cocktail name, photo, or PDF, it matches against real inventory and drafts a recipe with costs.

Curious how Crafti can elevate your operations? Here are examples of real questions operators have asked, to spark ideas on how you might gain insights.

Cost & Recipe Diagnostics

  • Why does our chocolate cake cost twice as much at this location compared to others?
  • Which signature cocktail has the highest profit margin?
  • Which limited-time offers generated the most profit over the past 12 months?

Cross-Location Analysis

  • Compared to 2024, which week in 2025 had our highest sales?
  • Comparing all locations over the last 30 days, which ones are underperforming?
  • Rank locations by sales from limited-time offers.

Operational Insights

  • Where is our biggest source of waste during this audit period?
  • How much Patrón Reposado have we sold since April?
  • What is driving differences in server performance, such as upsells or table turns?

Menu Engineering

  • Which menu items have the highest sales but the lowest profit margins?
  • Where can we raise prices with the least impact on sales?
  • Which menu items have seen the biggest sales decline over the past six months?

Labor & Scheduling

  • Which shifts have the highest labor cost as a percentage of sales?
  • How would adjusting staffing on specific shifts last month have impacted labor costs?
  • Which dates and times offer the best opportunity to reduce labor costs without impacting service?

Why Crafti Is Smarter Than a Generic LLM

While tech-savvy restaurant managers may have tried generic models, results are often underwhelming because they don't understand the individual business. Crafti is built differently from a general chatbot—it's a specialized assistant: a frontier model wrapped with Craftable-specific tools, product knowledge, and guardrails.

"Because it runs on your real operational data—POS, labor, invoices, inventory—it doesn't reason abstractly about the restaurant industry. It answers based on your suppliers, your recipes, your variances," Sundberg said.

Trust Built Into Every Answer

AI is only valuable when operators trust the answers. Crafti shows the work behind every response.

This means when Crafti investigates, it queries the data directly and returns the underlying numbers and its calculations. So instead of a vague conclusion like "your food cost looks high," a GM gets the actual figures supporting the conclusion, such as "theoretical is 28.4%, actual is 31.9%, a 3.5 percentage point variance."

This transparency is reinforced by the platform's architecture: data access is read-only by default, so investigations never alter data. Additionally, because access is isolated by location at the database level, Crafti only sees what the logged-in user is permitted to view—for example, just one location, not all.

For the two actions Crafti can take—drafting recipes or purchase orders—every draft is previewed for GM approval, and every action is recorded in an audit log. "The point isn't to get you to trust AI blindly, but to give you verifiable numbers and a traceable path," Sundberg said.

He added that this is just the beginning, and Crafti will become sharper as the platform evolves. "More agentic actions, broader module coverage, and tighter integration with other Craftable features are all on the roadmap."

Are you ready to spend less time hunting for answers and more time making better decisions?Click here to learn more about Craftable。