Building the First Artist Discovery Workflow

Every system starts somewhere.

For us, the first workflow was not the most advanced one. It was not the most automated one. It was the first clear step in turning a messy real-world process into something structured, repeatable, and easier to manage.

That workflow was simple in purpose:

• discover a new artist

• collect the important details

• check whether the artist already exists

• save the record if it is new

• let the human stay in control

That last part matters a lot. The workflow was never meant to replace people. It was meant to support them.

Why this workflow came first

Before you can build anything bigger, you need a clean way to capture new people into the system.

In a creative business, new artists can come from many places:

• a show

• a recommendation

• an agent

• a direct message

• a casual conversation

• a note taken by a manager

That means the input is often incomplete, informal, and inconsistent.

So the first problem we had to solve was not “how do we automate everything?”

It was:

How do we make sure a new artist is entered correctly, without losing the human judgment that matters?

The basic flow

The workflow begins when a literary manager or staff member notices a new artist.

From there, the process moves through a few simple steps:

1. Discover the artist

   • Someone meets or hears about a new artist.

2. Collect the details

   • Name

   • Email

   • Artist type

   • Location or southwest status

   • Phone, if available

3. Open the Artist Engine

   • The user clicks Add New Artist.

4. Validate the form

   • The system checks required fields and formats.

5. Search for possible duplicates

   • The system looks for similar names or records.

6. Let the human review

   • If there is a likely match, the user decides what to do.

7. Save the new artist

   • If it is a new person, the system creates the record in the database.

That is the core of the workflow.

Why the human step is important

This first workflow is a good example of something we believe strongly:

automation should help decision-making, not remove it

A system can check spelling. It can compare names. It can highlight duplicates. It can structure the data.

But a person still knows things the software does not know yet:

• whether the artist is actually new

• whether the name is spelled differently on purpose

• whether an existing record should be updated instead of duplicated

• whether the artist should be linked to a different team or region

That is why the workflow includes a human review point.

The human is not there to fight the system. The human is there to make the final call.

What makes this workflow useful

Even though it is simple, this workflow solves a big problem: it keeps the database clean.

Without this kind of process, records get messy very quickly:

• duplicate artists

• missing contact info

• inconsistent artist types

• unclear status fields

• records that cannot be trusted later

By putting a small amount of structure around the intake process, the workflow helps the whole system stay usable.

That matters because once a bad record gets into a database, it can cause problems in many other places later.

The role of AI in the future

At the moment, the first workflow is mostly about structured human input.

But this is also where AI could help later.

AI could support this workflow by:

• reading rough notes and turning them into clean fields

• suggesting duplicate matches more intelligently

• flagging missing information

• helping normalize artist names and details

• making the intake faster without removing human review

That makes the first workflow a strong foundation for future automation.

It is a good place to start because it already has clear decision points where AI can assist, but a human can still stay in charge.

What this workflow taught us

The most important lesson from the first workflow was not technical.

It was about design.

We learned that a good workflow should:

• start with a real human action

• ask for only the important information

• check for obvious errors early

• keep the database clean

• leave room for judgment

• make the next step easier

That mindset shaped the rest of the system.

Once we had a clear way to add new artists, it became much easier to think about other workflows too:

• work submissions

• meeting packs

• verdict logging

• duplicate review

• archiving

The first workflow became the base layer.

Final thoughts

The first artist discovery workflow was simple, but that was the point.

It turned a vague, manual process into something visible and dependable.

It gave the team a way to:

• capture new artists properly

• avoid duplicate records

• trust the data more

• keep humans in control

• prepare the system for smarter automation later

Sometimes the most important workflow is not the most exciting one.

It is the one that makes everything else possible.