Skip to main content

What Konekti AI can do

K-AI helps you across the three activities you already work in: modeling, analysis, and debugging. This page maps out what K-AI can take on and where its limits are. For how to prompt K-AI so you get the best results, see Getting the best from Konekti AI.

Modeling​

K-AI helps you turn raw data into a solid model:

  • Explore your sources. Ask what tables and columns you have connected and what's in them, without leaving the conversation.
  • Build and refine your model. Based on your data and what you want to analyze, K-AI can suggest and create your assets (objects & activities), relations, and processes.

You can ask things like "What's in my connected sources?" or "Suggest objects and activities for my order process."

Analysis​

K-AI answers questions about your processes in plain language, so you don't have to build a query yourself:

  • One case or all of them. Ask about a single order's journey, or about patterns across every case: durations, bottlenecks, variants, trends, and more.
  • Grounded in your data. Every answer is based on your real data, and you can trace it back to where it came from.

You can ask things like "Which step takes the longest in the order process?" or "How many invoices are linked to order 1043?"

Debugging​

When something looks off, K-AI helps you find out why:

  • Diagnose the problem. K-AI tracks down issues like a missing relation between objects, an order with no linked invoice, or a number that doesn't add up, and explains the likely cause.
  • Point you to a fix. When the fix is a change to your model, K-AI proposes it for you to review.

You can ask things like "Why do some orders have no linked invoice?" instead of "Something is wrong, just fix it."

K-AI proposes, you decide​

K-AI never changes your model on its own. When it wants to add or fix something, it shows you a proposal that you accept or decline, so you stay in control of what ends up in your workspace. See Proposals in K-AI on how K-AI works. to learn how K-AI gives you proposals and what you can do with it.

What K-AI can't do​

Knowing the limits helps you ask sharper questions. K-AI does not:

  • Look across workspaces: K-AI works within one workspace at a time.
  • Act outside Konekti: It can't send emails, create tickets, or trigger other systems.
  • Analyze what isn't modeled: If something isn't part of your model yet, K-AI can't analyze it.
  • Reprocess your data for you: K-AI helps you build and fix your model. You stay in control of when the data behind it is processed.

When you ask for something outside these limits, K-AI won't just stop. It explains what's missing and points you toward the next step.

tip

Knowing what K-AI can do is only half of it. How you ask is the other half. See Getting the best from Konekti AI.

Learn more about what how K-AI reaches an answer, and how to read them on the next page.