How AI is accelerating automotive technical training
- 7 minutes read
Imagine if you could find the answer to a complex technical question instantly, instead of trawling through workshop manuals or databases.
This sort of search already exists in the form of generative AI applications such as ChatGPT, which scrape the web to give users instant answers to almost any prompt.
But these consumer-focused AI solutions aren’t very adept at handling complex data. They struggle with unstructured documents, picture interpretation and are limited in volume. Furthermore, they tend to hallucinate if they don’t know the answer, and they’re only as reliable as the data source underpinning it.
That’s why, in Parts & Service departments where accuracy is paramount, these off-the-shelf gen-AI solutions haven’t proven to be robust enough to help technicians and customer relations teams streamline their work.
Intellegam – reinventing knowledge access
But that is changing fast. Intellegam is a search-focussed AI solution that turns complex organizational data into digital, easily accessible knowledge.
The company, which started life in 2023, was recently acquired by Infomedia, and is designed to provide teams with fast answers to reduce training time and improve accuracy.
“The user seeks to access insights hidden within the company’s data,” said Marc Gehring, CTO at Intellegam. “Our solution streamlines that process and enables users to spend less time searching and more time acting on what matters.”
For instance, Intellegam can improve tech efficiency by merging diagnostics and repair procedures into vehicle inspection workflows. If a technician asks, for instance, “where is the location of the oxygen sensor?” Intellegam will instantly find the answer for that specific vehicle, and can even pull up a step-by-step repair process. It even provides diagnosis procedures and interpretation of DTC codes or pulls up the right wiring diagram.
But Intellegam’s power extends beyond the Parts & Service department’s workshop – it can be a valuable customer communication tool, too.
Intellegam’s technology helps call center agents to find the answer to almost any question using natural language, ready to use in an email response or in a conversation. Its channel agnostic nature enables the customers to also interact via chat or voice.
It can even be used to enhance service advisor efficiency by analyzing the vehicle health check report and writing customer emails recommending further repairs and potential up-sell opportunities.
Powerful search, enhanced with AI
So how does it work? Intellegam collects OEM data, structures it, integrates it with powerful search capability, enhances it with AI, and delivers it through the preferred channels.
Intellegam’s proprietary intelligent search algorithm can fully automated scrape complex and extremely large data sets. That can be anything from Powerpoint presentations, images and video, Confluence pages, tooling documentation, part numbers, schematics, internal FAQs and user manuals.
Once this information is in the system, Intellegam structures the data, and presents it in an intuitive language agnostic user interface. It’s a bit like a supercharged version of Ctrl+f.
“We just need the information, then we can pull the data into our secure, cloud and make sense of it by turning it into small, contextual snippets,” said Hannes Burrichter, CIO at Intellegam[MZ1] .
Intellegam has developed its own algorithm for semantic information retrieval – a core part of its intellectual property – leveraging multiple AI tools to master steps where most other companies struggle.
All data is GDPR compliant and processed with the highest data protection standards in accordance with EU guidelines.
Save technicians’ time
Intellegam makes available in an instant information that is otherwise often buried in comprehensive training materials or databases, or requires a lot of manual searching or cross-referencing.
For Parts & Service departments it means, for instance, technicians can diagnose and repair faults by finding the solution quickly. Or customer relations team can promptly respond to customer queries. And because Intellegam is fed by OEM data, it’s always accurate and reliable.
“The digitization of organizational knowledge, the implementation of intelligent assistance systems, and the establishment of a structured knowledge base are critical strategies to ensure the continuity of business operations and sustain innovation capacity,” said Gehring.
Getting the information into the system doesn’t require much effort either, or the creation of large data lakes. OEMs and dealers can simply share data.
Intellegam can fully automate the process of integrating information into the system without any manual effort, delivering a significant efficiency advantage.
Automated multi-language translation
Over the top of this sits a large language model (LLM), which users can prompt for instantaneous answers.
Intellegam’s solution is LLM-agnostic and can be seamlessly integrated with any large language model, including ChatGPT, Claude, DeepSeek, or others, ensuring maximum flexibility for clients.
Using the LLM’s API, Intellegam can work with whatever solution the customer desires, and it can be white-labelled and incorporated as an AI assistant in websites, workflows, email, sharepoints, knowledge management platforms or data servers.
In addition, the information can be outputted in whatever format you want, whether that’s a simple text answer, an image, or a table comparing data, in multiple languages.
How Nio boosted efficiency
Intellegam has been working with Chinese OEM Nio to power its knowledge management database for its European call center in five different languages. To date, it has ingested more than 10,000 pages of information.
“As a premium brand and ‘User Enterprise’, NIO has high standards when it comes to user support. With Intellegam’s solution, we are able to respond to our clients request faster at our high standards, which increases customer satisfaction and efficiency,” said Zhengke Zhou, business manager at Nio Europe.
The LLM can handle all languages. If a customer calls in, the staff can type in a question and get an instant answer, even if the source of that information is in the native language.
And if the call handler or the technician wants to verify if the information is accurate, they can simply click on the source, and it will take them straight to the exact place in the document where the information has been extracted.
The system makes it very transparent. Users can jump to the source document, to the right page and section so they can see it is correct.
One of the main benefits for Nio, explains Gehring, has been a reduction in training time. “There’s a lot of staff turnover in call centers, and so there’s a constant need to retrain staff because the deeper you train, the better the customer experience.”
However, Nio reduced training time by 50% using Intellegam. This meant it could refocus its training on customer interaction and communication rather than knowledge retention and sharing information that the staff might forget.
Nio also saw an increase in first time resolution rate, and a 68% reduction in the average time it takes to find a solution.
In call centers they reduced the search time by almost 70% from 10 minutes down to three minutes.
For another Chinese OEM, Intellegam supported technicians in getting instructions on diagnosis steps and repair procedures in a natural dialog with immediate responses.
The OEM saw a reduction in “tec-line” tickets, an increase in technicians’ daily efficiency and the immediate available of manuals or bulletins for technicians use, with translation handled automatically by the LLM.
“Intellegam’s technology impressed me with its ability to automatically digest thousands of pages of repair instructions, pictures and wiring diagrams and to come up with the exact diagnosis and step-by-step diagnosis and repair instructions for the technicians,” said the client’s head of technology.
“Despite the fact the information fed into the system was challenging in terms of structure and content (being wiring diagrams and pictures), the robustness of the data processing was mind blowing and clearly is not comparable to standard solutions as the likes of ChatGPT. The solution works with perfect accuracy, quality and speed after an implementation time of eight weeks.”
“Strategic knowledge management is no longer optional; it is essential for business success,” said Gehring. “Organizations that invest in systematic knowledge retention and sharing today will secure a significant competitive advantage for the future.”