# A likely diagnosis, with the repairs that back it up.

Canonical: https://www.aiventic.ai/platform/service-reasoning

> Service Reasoning weighs your past repairs and the documentation to rank likely causes, flag fixes that already failed, and show the evidence for each.

Layer 3 · Service Reasoning

Aiventic compares the job to every similar repair you’ve done and the documentation for the unit, then ranks what’s likely wrong and shows why.

[Book a demo ](https://www.aiventic.ai/schedule-a-demo)[See the whole platform](https://www.aiventic.ai/platform)

1. [L4 Service Action What should happen next?](https://www.aiventic.ai/platform/service-action)
2. [L3 Service Reasoning What does this situation likely mean?](https://www.aiventic.ai/platform/service-reasoning)
3. [L2 Service Knowledge What does the documentation say?](https://www.aiventic.ai/platform/service-knowledge)
4. [L1 Service Memory What happened?](https://www.aiventic.ai/platform/service-memory)

## “What does this situation likely mean?” The question Service Reasoning answers.

What goes in

* The request: model, symptom, what was tried
* Unit history from Service Memory
* Similar repairs and their outcomes
* Manuals and bulletins from Service Knowledge
* Recent fixes on the unit that didn’t hold

What comes out

* Ranked likely causes
* Evidence behind each cause
* Probable parts
* A diagnostic path to follow
* Repeat-failure flags
* Odds a fix holds on the first visit

## What it does. Inside Service Reasoning.

* 01  
### Finds the repairs that look like this one  
Same model, same symptom, same history. It finds the jobs that actually match, then checks which fix held.
* 02  
### Rules out what already failed  
If a part was swapped on this unit and the problem came back, that fix drops down the list instead of getting tried again.
* 03  
### Shows its evidence  
Every cause comes with the past jobs and documents behind it, so a tech or dispatcher can check the reasoning, not just trust it.
* 04  
### Spots repeat failures  
When a unit keeps coming back for the same symptom, Aiventic flags it before another truck rolls with the same part.

## One job, one layer. The Hernandez oven call, through this layer.

How Service Reasoning reads the Hernandez job

Demo workspace

Similar repairs

47 on the same model and symptom

Ruled out

Igniter replaced 21 days ago, didn’t hold

Documentation

OEM bulletin + service manual p.38

Likely cause

Ignition control failure, 86%

Fix that holds

Control swap held on first visit 89% of the time

## Where you’ll see it. Solutions built on Service Reasoning.

* [Similar Repair Intelligence See every time you’ve fixed this model and symptom, and what held Learn more ](https://www.aiventic.ai/solutions/similar-repair-intelligence)
* [Diagnostic Guidance Ranked likely causes with the evidence, and the next check to make Learn more ](https://www.aiventic.ai/solutions/diagnostic-guidance)
* [Reduce Callbacks Stop paying twice for the same repair by fixing it right the first time Learn more ](https://www.aiventic.ai/solutions/reduce-callbacks)
* [Increase First-Call Completes Arrive with the right part and a likely diagnosis, and finish on visit one Learn more ](https://www.aiventic.ai/solutions/increase-first-call-completes)

## Questions. Straight answers.

Is this a generic AI guess?

No. The reasoning starts from your own closed jobs on the same model and symptom, then the documentation for the unit. Every recommendation shows the evidence behind it.

What if the likely cause is wrong?

The tech sees the ranked causes and the diagnostic path, not a single answer. Whatever actually fixed it flows back in, and the next call on that model reasons from it.

Can agents use the reasoning directly?

Yes. ChatGPT, Claude, and FSM agents can ask for similar repairs, likely causes, and probable parts over MCP and the Aiventic API.

## Stop solving the same repair for the first time.

Bring a few hundred closed jobs. We’ll show you what your company already knows.

[Book a demo ](https://www.aiventic.ai/schedule-a-demo)

---

*Markdown representation of [https://www.aiventic.ai](https://www.aiventic.ai). Append `.md` to any page URL, or send `Accept: text/markdown`, for its markdown. Full site index: [/llms.txt](https://www.aiventic.ai/llms.txt) · [/sitemap.xml](https://www.aiventic.ai/sitemap.xml)*