Create a Custom AI ChatBot using your own data

ByJorge Pereira
Tags: ,

Part of: AI Learning Series Here

Using AI can save tons of time. Increases your productivity exponentially. However public AI Bots tools do not allow you to use your own data at scale, being limited to uploading a single document (typically a limited sized txt or PDF)

If you are looking to take advantage of your own data to help users chat / interact in a natural language to get answers, then you can create your own chatbot without having to create your own language model (LLM). It is a good way to understand how AI can be customized for you own use case.

There are fairly good tools that do this for you already, but in this document I am focusing on explaining some of the basics instead. For a list of tools: go HERE ( Gradient AI, Bubble.io, Flowise)

The steps are fairly straight forward:

  1. Choose the core LLM you want to use
    • (you can use the OpenAI LLM API)
  2. Add your own data source ( Pick or create a location with your source files)
  3. Once uploaded you will need to “index” your documents.
    • (Please note that your data is NOT going back to the LLM (Model). You are not training that model, you are using that model to index your data!)
  4. Fine Tune the model which is basically adjusting the tone, limit responses, and things like that.
  5. Create the end-user front-end
    • You can use HTML/WebApp or integrate into your own application using API calls.

Best explanation videos I have seen on this topic are:

Fine Tuning:

In machine learning, fine-tuning is the process of optimizing a model for a specific dataset. This can be done by adjusting the model’s parameters, such as the learning rate or the number of hidden layers. 

A few examples from YouTube:

BUilding using tools:

Accomplished Information Technology services professional with over 25 years of experience performing technical pre-sales, solution selling, public speaking / presentations, consulting, project delivery and program management. Over the past few years, Jorge has focused on customers solutions in the areas of: Artificial Intelligence, Generative AI, Workforce transformation, End-User-Computing (EUC) Lifecycle Management, Modern Device Management, and Cybersecurity.

Disclaimer:  I work for Dell Technology Services as a Workforce Transformation Solutions Principal.    It is my passion to help guide organizations through the current technology transition specifically as it relates to Workforce Transformation.  Visit Dell Technologies site for more information.  Opinions are my own and not the views of my employer.

Similar Posts

Windows 11 is here
|

Windows 11 is here

ByJorge Pereira

Update 7/10/2021 Minimum hardware requirements further explained – A challenge for many! To run Windows 11, devices must have an Intel Core processor from at least 2017, or AMD Zen 2 processors from 2019 onward. Video of installation on a Hyper-V virtual Machine (my own test) Microsoft releases second public review – with ONLY minor…