5 steps for building a ‘no code’ chatbot in under 5 minutes
On March 21st, the NLP-CoP’s Sandbox Working Group hosted a webinar on creating custom chatbots using paid (ChatGPT Plus) and free (Hugging Face) no-code platforms. Zach Tilton and I guided participants through the development of simple chatbots leveraging OpenAI’s GPT and Hugging Face Assistants. The webinar aimed to offer ways to enhance Monitoring, Evaluation, Research, and Learning (MERL) activities by integrating AI-driven chat technologies.
Participants engaged in a practical demonstration, gaining a guided experience in chatbot construction. We also explored how chatbots could facilitate data collection, data analysis, and streamline knowledge management. The lively discussion allowed attendees to share their questions and get a better understanding of how these tools can be leveraged.
This webinar was targeted at novices and those who have not yet experimented with these tools.
Below I’ve created a series of screenshots to walk you through the steps we shared at the meeting. Happy building!
Steps for building a quick chatbot!
Step 1: Either purchase a ChatGPT Plus account (USD 20 per month) or create a free account on Hugging Face.


Step 2: Click on ‘Explore GPTs’ on ChatGPT or ‘Hugging Chat’ then ‘Assistants’



Step 3: On ChatGPT Plus click on ‘Create’ and on Hugging Face click on ‘Create new assistant’


Step 4: In ChatGPT Plus, describe to the GPT Builder what you want the GPT to do. You can also add custom instructions under the ‘Configure’ tab. In Hugging Face, add what you want the assistant to do under the ‘Instructions (System Prompt)̈’ field.


Step 5: Start using your new AI Assistant! You can continue to test and tweak it as you go. Have fun!


You might also like
-
Keeping the Human in the loop: Balancing AI capabilities with Human Judgement in Evaluation
-
Join us in Bangkok on December 1st: Free workshop on user research and digital design
-
Opening the Gates: A Risk Taxonomy for Inclusive AI Ag Advisory
-
Join us on October 28 for a discussion on AI and SBC program design with Elizabeth Long
