<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>3322</id>
    <completedYear>2025</completedYear>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst>695</pageFirst>
    <pageLast>706</pageLast>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>conferenceobject</type>
    <publisherName>SCITEPRESS</publisherName>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>1</belongsToBibliography>
    <completedDate>2025-04-25</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Towards Personal Assistants for Energy Processes Based on Locally Deployed LLMs</title>
    <abstract language="eng">This paper presents a coaching assistant for network operator processes based on a Retrieval-Augmented Gen-&#13;
eration (RAG) system leveraging open-source Large Language Models (LLMs) as well as Embedding Models.&#13;
The system addresses challenges in employee onboarding and training, particularly in the context of increased&#13;
customer contact due to more complex and extensive processes. Our approach incorporates domain-specific&#13;
knowledge bases to generate precise, context-aware recommendations while mitigating LLM hallucination.&#13;
We introduce our systems architecture to run all components on-premise in an our own datacenter, ensuring&#13;
data security and process knowledge control. We also describe requirements for underlying knowledge doc-&#13;
uments and their impact on assistant answer quality. Our system aims to improve onboarding accuracy and&#13;
speed while reducing senior employee workload.&#13;
The results of our study show that realizing a coaching assistant for German network operators is reasonable,&#13;
when addressing performance, correctness, integration and locality. However current results regarding accu-&#13;
racy do not yet meet the requirements for productive use.</abstract>
    <parentTitle language="eng">Proceedings of the 17th International Conference on Agents and Artificial Intelligence (ICAART 2025) , 2025, Porto, Portugal</parentTitle>
    <identifier type="doi">10.5220/0013175600003890</identifier>
    <enrichment key="PeerReviewNachweis">ja</enrichment>
    <enrichment key="opus.source">publish</enrichment>
    <licence>Creative Commons - CC BY-NC-ND - Namensnennung - Nicht kommerziell - Keine Bearbeitungen 4.0</licence>
    <author>Maximilian Orlowski</author>
    <author>Emilia Knauff</author>
    <author>Florian Marquardt</author>
    <collection role="open_access" number="">open_access</collection>
    <collection role="institutes" number="">Fachbereich Informatik und Medien</collection>
    <collection role="Hochschulbibliografie" number="1">Hochschulbibliografie</collection>
  </doc>
</export-example>
