Hugging Face moves into no-code workflows

This is a move from serving developers as infrastructure to owning the actual business user workflow. HuggingChat turns Hugging Face’s model hub into a simple app surface where someone can create a custom chatbot without writing Python, which starts to overlap with the kind of lightweight internal tools and workflow apps that product managers often patch together in Airtable or Zapier.

• Zapier’s core value is taking a plain language request, turning it into app actions across 6,000 plus services, then hiding API complexity, auth, and field mapping. That is exactly the user behavior Hugging Face can go after if it wraps models, tools, and deployment in a similarly simple interface.
  3 sacra 4 zapier
• Airtable is pushing the same buyer in a different direction. Cobuilder lets a non technical user describe a business process and get a working app in seconds. The overlap is the person who knows the workflow problem well, but does not want to build from scratch in code.
  5 airtable 6 airtable
• The bigger prize is seat expansion inside enterprises. Hugging Face already had an estimated $70M ARR by the end of 2023, with most revenue tied to managed enterprise usage and services. Low code app surfaces create a path from platform adoption by ML teams to daily use by product, ops, and analytics teams.
  1 sacra 2 sacra 7 sacra

The next step is a broader no code layer on top of open models, datasets, tools, and hosted inference. If Hugging Face keeps making model powered apps easier to create than a Zap or an Airtable base, it can become the default place where non developers build AI native internal software, not just the place where developers download models.