> For the complete documentation index, see [llms.txt](https://autodoc-1.gitbook.io/autodoc-product-doc/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://autodoc-1.gitbook.io/autodoc-product-doc/core-feature-1-ai-diagnostic-agent.md).

# Core Feature 1: AI Diagnostic Agent

**What Problem Does This Solve?**

The automotive repair industry has long suffered from a fundamental structural problem: a severe information asymmetry between car owners and repair shops. Most car owners have little to no mechanical knowledge. When a dashboard warning light comes on, an unusual noise appears, or something feels off, they have no choice but to walk into a shop blind — entirely dependent on the mechanic's judgment. This information gap is the root cause of over-servicing, inflated bills, and the general distrust consumers feel toward repair shops.

AutoDoc's AI Diagnostic Agent was built specifically to break down this wall. Its true purpose is to serve as a transparent bridge between car owners and repair shops: before a car owner ever sets foot in a garage, they can use AI-powered conversation to understand what might be wrong with their vehicle, get a realistic estimate of repair costs, and find out which nearby shops are trustworthy. The car owner is no longer the uninformed party — and repair shops benefit too, as transparency builds the kind of trust that drives long-term customer relationships.

In one sentence: ***The AI Diagnostic Agent is not just a chatbot — it's every car owner's personal advisor before they enter a repair shop, ensuring they walk in with equal footing and informed expectations.***

**Interface Layout**

Visit <https://www.autodoc.network/chat> to access the AutoCar AI Agent. [3](https://www.autodoc.network/chat) The interface is organized into three zones:

* **Left sidebar:** Displays your diagnosis history and a "+ New Diagnosis" button to start a fresh conversation
* **Top toolbar:** A vehicle selector dropdown (e.g., CHEVROLET SILVERADO 1500) to switch between registered vehicles
* **Main chat area:** The interactive multi-turn conversation space between you and the AI

<figure><img src="https://2919185632-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBH85joWTAIYkzAWl5P8H%2Fuploads%2FW4puKnOrMLItDFh7R25N%2Fimage.png?alt=media&amp;token=44869ff1-06fc-4000-9530-8535c1ae2c0e" alt=""><figcaption></figcaption></figure>

**Three Input Methods**

The AI Agent accepts vehicle problem descriptions in three ways:

<table data-header-hidden><thead><tr><th width="212.95703125"></th><th width="401.578125"></th><th></th></tr></thead><tbody><tr><td><strong>Input Method</strong></td><td><strong>Description</strong></td><td><strong>Status</strong></td></tr><tr><td>Text description</td><td>Type your issue directly into the chat input</td><td>Live</td></tr><tr><td>Upload diagnostic report</td><td>Submit a report exported from an OBD scanner or third-party diagnostic tool</td><td>Live</td></tr><tr><td>Dashboard photo</td><td>Take a photo of your dashboard warning lights for visual AI analysis</td><td>Live</td></tr><tr><td>Connect OBD device</td><td>Direct real-time connection to the vehicle's on-board diagnostic port</td><td>Coming Soon</td></tr></tbody></table>

**Step-by-Step Diagnosis Flow**

The AI Agent uses a structured, multi-turn diagnostic conversation modeled after how a professional mechanic would triage a problem:

**Step 1 — Describe the symptom**

Describe the issue in plain language, for example: *"My engine warning light is on, what should I do?"* or *"My AC is running but the car is still very hot."*

<figure><img src="https://2919185632-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBH85joWTAIYkzAWl5P8H%2Fuploads%2FsEzqkHsyeWQleiO69Vit%2Fimage.png?alt=media&amp;token=444c2862-a6f0-4ee9-9b6d-816ba8d730bf" alt=""><figcaption></figcaption></figure>

**Step 2 — AI asks clarifying questions**

Based on your initial description, the AI will ask one or two targeted follow-up questions to narrow down the possible causes. For example, in response to an engine warning light:*"Besides the engine warning light, have you noticed any of the following?"*

1. Engine vibration or unstable idle
2. Weak acceleration or hesitation
3. Unusual smell (gasoline or burning odor)
4. No other obvious symptoms
5. Other abnormalities — please describe

<figure><img src="https://2919185632-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBH85joWTAIYkzAWl5P8H%2Fuploads%2F2QLg9kyyQxK7IxbSw2Tk%2Fimage.png?alt=media&amp;token=b25d6abf-8953-438b-8b01-7c473c9c3bf8" alt=""><figcaption></figcaption></figure>

**Step 3 — Diagnosis report generated**\
Based on your selections, the AI produces a structured DIAGNOSIS panel listing detected issues with their severity levels (High / Medium / Low), accompanied by professional explanations and recommended next steps.

<figure><img src="https://2919185632-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBH85joWTAIYkzAWl5P8H%2Fuploads%2F6FtOO4wcrd6ZdALLu9kM%2Fimage.png?alt=media&amp;token=5849e9f9-6003-423d-9dbf-205f4d08a9c8" alt=""><figcaption></figcaption></figure>

**Step 4 — Repair estimate and shop recommendation**

After the diagnosis, you can ask follow-up questions such as:

* "Give me a repair estimate" — the AI provides a price range broken down by labor and parts costs
* "Which repair shops near me do you recommend?" — the AI surfaces nearby partnered shops based on your location
* "Is it safe to keep driving?" — the AI assesses the immediate safety risk of the diagnosed issue

As a reference, here is an example repair estimate for an AC fault:

<figure><img src="https://2919185632-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FBH85joWTAIYkzAWl5P8H%2Fuploads%2FBcl0CuStXPBRfSlPWqK8%2Fimage.png?alt=media&amp;token=043c0636-449f-45be-b2c9-e8b81ea440d2" alt=""><figcaption></figcaption></figure>