Calendario Preguntas y Respuestas
Calendario y Tarifas Preguntas y Respuestas
Servicios Descripción de Servicios Preguntas y Respuestas
C1Fundamentos C5Gobernanza C10Chatbots C3Analítica C2Alfabetización C12Integración C11Accesibilidad C8Marketing

FAQ ChatBots and Basic Agents for Libraries (open source · minimum viable prototype)

Build your own library ChatBot: from scratch to a functional prototype with RAG (Retrieval-Augmented Generation) and security controls.

Your technical team builds a FAQ ChatBot with RAG (Retrieval-Augmented Generation) using open-source tools, implements security controls, evaluates it with real users and documents its ongoing operation.

Problems it solves

  • Repetitive queries that consume library staff time.
  • Lack of 24/7 support for frequently asked questions.
  • Dependence on costly vendors for conversational solutions.
  • No criteria to evaluate the quality and safety of a ChatBot.

Results achieved

  • Minimum viable FAQ ChatBot prototype with RAG (Retrieval-Augmented Generation).
  • Security controls implemented: anti-hallucination, scope limits, traceability.
  • Prototype evaluation with quality and user metrics.
  • Operational maintenance and continuous improvement guide.

Course structure

Types of ChatBots and when to use each. RAG (Retrieval-Augmented Generation) architecture: how it works and why it is more reliable. Available open-source tools. Defining the scope of the prototype.

Preparing and structuring documents: FAQs, policies, service guides. Chunking, embeddings and vectorization. Workshop: loading the initial knowledge base for the prototype.

Development environment setup. Implementing the RAG pipeline: retrieval + generation. Guided workshop: first functional ChatBot conversations.

Conversational flows: handoff to human, boundary messages, handling out-of-scope questions. Workshop: testing with real library cases.

Anti-hallucination policy: how to make the bot cite sources or declare uncertainty. Quality evaluation: resolution rate, information gaps, satisfaction. Adjustments based on test results.

Web deployment options and other channels. Basic analytics panel. Collective creation of the operational guide: updating content, reviewing quality, managing improvements. Close and continuity plan.

Frequently asked questions

Yes. This is an intermediate-level course involving technical tools. Basic familiarity with Python or the command line is recommended. The instructor guides every step.

Open-source tools selected based on the group's context. The approach prioritizes self-hostable, no-license-cost solutions.

Minimum viable ChatBot prototype with RAG, implemented security controls, documented evaluation and operational maintenance guide.

At least 10 hours to complete workshops and adapt the prototype to your institutional context. This is the most intensive course in the portfolio.

Yes, as a functional prototype ready for a pilot. Your library team is trained to operate and improve it.

← All courses ← C5 Next: C3 →

Your Library. Your Evolution.

Artificial Intelligence critical and responsible for libraries

View all courses AI Implementation