Skip to content

4.1: Understanding Action Sequences

You've built a chatbot. You've added knowledge. You've customized its personality. Now let's talk about how all these pieces work together — and most importantly, how they happen automatically without you lifting a finger!

The big idea: Your agent follows a sequence of actions, just like following a recipe. Each step happens automatically when it's ready, in the right order, every single time.

💡 What You'll Learn

By the end of this section, you'll understand:

  • ✅ Why action sequence matters (hint: you can't frost a cake before baking it!)
  • ✅ How your agent processes each message step-by-step
  • ✅ What "waiting" really means (and why it's a good thing)
  • ✅ How automatic execution saves you time and prevents errors

4.1.1: The Cooking Analogy

Let's start with something familiar: baking a cake.

When you bake a cake, you can't just do things in any random order. Try adding frosting before baking, and you'll have a melted mess! Some steps must happen in sequence:

🎂

Baking a Cake: Steps Must Happen in Order

Just like following a recipe, your automation needs to do things in the right order. You can't frost a cake before it's baked!

1

Mix ingredients

Combine flour, sugar, eggs, butter

2

Pour into pan

Transfer batter to baking pan

3

Bake at 350°F for 30 minutes

Wait for it to cook through

4

Let cool completely

Must cool before frosting!

5

Add frosting

Only possible after cooling! ✨

Can't skip steps: If you try to add frosting before baking, it will melt! Each step depends on the previous one finishing.

Right order = success: When you follow the sequence, you get a perfect cake!

The same is true for your AI agent. It follows a sequence of actions:

  1. Receive the customer's question
  2. Search your knowledge base (if needed)
  3. Generate a helpful response
  4. Send it back to the customer

Just like with baking, you can't skip ahead. The agent can't generate a response before it knows what the question is!

💡 💡 Think of It Like This

Your automation is like an assembly line.

Each column is a workstation. Each row is an item moving through the line. The item can't move to the next workstation until the current one finishes its work.

Example:

  • Station 1: Receive customer question → ✅ Done
  • Station 2: Search knowledge → Waiting for Station 1 to finish...
  • Station 3: Generate response → Waiting for Station 2 to finish...
  • Station 4: Send to customer → Waiting for Station 3 to finish...

Everything flows smoothly, one step at a time!


4.1.2: How Your Agent Processes Messages

Let's watch your chatbot in action and see exactly what happens, step by step:

How Your Agent Processes Messages

Here's the sequence your chatbot follows for every customer message:

1

📥 Receive Message

Customer question arrives in your table

"What are your business hours?"
2

🔍 Search Knowledge (if needed)

AI looks through your documents for relevant information

Found: "Business Hours: Monday-Friday 9am-5pm EST"
3

✨ Generate Response

AI crafts a helpful answer using the knowledge it found

"We're open Monday through Friday, 9am to 5pm Eastern Time. Happy to help you during those hours!"
4

📤 Send to Customer

Response goes back to the customer instantly

✓ Message delivered!

The key insight: Each step waits for the previous one to complete. You can't generate a response before receiving the message, and you can't send a response before generating it!

What's happening here?

Each step waits for the previous step to complete:

  • Step 2 can't start until Step 1 delivers the message
  • Step 3 can't start until Step 2 finds the information
  • Step 4 can't start until Step 3 creates the response

This is called a dependency chain. Each action depends on the previous one finishing.

💡 Why This Is Powerful

Automatic ordering means you never have to worry about:

  • Running steps out of sequence
  • Forgetting to run a step
  • Using data before it's ready
  • Manually clicking "run" for each action

It just works — every single time!


4.1.3: Understanding "Waiting"

When you look at your table, you might see cells with different statuses. Let's understand what each means:

Watching Actions Progress

Each cell shows its current status. Watch as it progresses automatically:

Customer QuestionAI ResponseSend to Customer
What are your business hours?
Completed
Ready
Waiting...
Ready
Ready← Current

Waiting for input data

Queued

Scheduled to run

Running

Processing now

Completed

Finished successfully

Automatic execution: Once all requirements are met, the action happens automatically. You don't need to click "run" or manually trigger anything!

What do these statuses mean?

Ready

The cell is waiting for its input data. Once the data arrives (from a previous column or from you), it automatically moves to the next state.

Think of it like: A chef standing by, ready to cook as soon as the ingredients arrive.

Queued

The cell has everything it needs and is scheduled to run. It's in line, waiting its turn.

Think of it like: Your order at a coffee shop — it's been placed, now waiting for the barista to make it.

Running

The action is actively processing right now! This might take a few seconds (for a quick lookup) or longer (for an AI-generated response).

Think of it like: The barista is actively making your coffee.

Completed

All done! The result is ready. The next column can now use this data.

Think of it like: Your coffee is ready — time to enjoy!

💡 Important: 'Waiting' Is Good!

Seeing "Ready" or "Queued" doesn't mean something is broken. It means the system is correctly waiting for what it needs.

Example:

  • Your "AI Response" column shows "Ready" → Perfect! It's waiting for the customer question to arrive
  • Customer types a question → "AI Response" changes to "Queued" → Then "Running" → Then "Completed"

This is exactly how it should work!


4.1.4: The Magic of Automatic Execution

Here's where Uniki really shines. You don't have to click "run" for each step.

Let's compare:

Automatic Execution: No Clicking Required!

❌ Other Tools (Manual)

1. Type customer question
2. Search knowledge base
3. Generate AI response
4. Send to customer
Tedious! You have to click "Run" for every single step.

✅ Uniki (Automatic)

Type customer question
→ Automatically searches knowledge
Search completes
→ Automatically generates response
Response generated
→ Automatically sends to customer
Done! 🎉
Effortless! Just add the question, and everything else happens automatically.

Why Automatic Execution Matters:

Faster

No waiting for you to click buttons

🎯
Reliable

Never forget to run a step

🌙
Works 24/7

Runs even when you're asleep

📈
Scalable

Handle hundreds of requests automatically

Bottom line: Set it up once, and it runs automatically forever!

The difference is huge:

Other tools: You set up your automation, but then you have to manually trigger each step. Every. Single. Time.

  • New lead comes in → Click "Run Step 1"
  • Step 1 finishes → Click "Run Step 2"
  • Step 2 finishes → Click "Run Step 3"
  • And so on...

Uniki: You set up your automation once. Then it runs automatically forever.

  • New lead comes in → Everything happens automatically
  • You just check back later to see the results
  • Or don't check at all — it's working 24/7 without you!

💡 Real-World Impact

Imagine you get 100 customer questions per day.

Without automatic execution:

  • 100 questions × 4 steps each = 400 manual clicks
  • At 5 seconds per click = 33 minutes of repetitive clicking
  • Every. Single. Day.

With automatic execution:

  • 100 questions × 0 clicks = 0 minutes
  • You set it up once, it runs forever
  • You can focus on important work while automation handles the repetitive stuff!

4.1.5: Putting It All Together

Let's review what you've learned:

1. Actions happen in sequence

  • Just like following a recipe, each step happens in order
  • You can't skip ahead or do things out of sequence
  • This prevents errors and ensures reliable results

2. Each action waits for what it needs

  • Columns wait for their input data to be ready
  • "Waiting" statuses (Ready, Queued) are normal and good
  • Once requirements are met, execution happens automatically

3. Automatic execution is the secret sauce

  • No clicking "run" for each step
  • Works 24/7, even when you're asleep
  • Scales effortlessly to hundreds or thousands of items
  • Saves hours of repetitive work

4. You stay in control

  • See exactly what's happening with status indicators
  • Understand why something is waiting
  • Troubleshoot easily when needed

💡 🎯 Try This!

Test automatic execution yourself:

  1. Open your chatbot table from earlier tutorials
  2. Add a new row with a customer question
  3. Watch the status badges change automatically:
    • Customer Question: Completed (you typed it)
    • AI Response: Ready → Queued → Running → Completed
    • (If you have a send column): Ready → Queued → Running → Completed

You didn't click anything! The system knew what to do and when to do it.

That's the power of action sequences!


What You've Learned

🎉 Congratulations! You now understand how your automations actually work under the hood.

Key takeaways:

  • Sequence matters — just like baking a cake, you need to do things in the right order
  • Dependencies create flow — each action waits for what it needs, then runs automatically
  • Statuses tell the story — Ready, Queued, Running, Completed show you exactly what's happening
  • Automation runs itself — once set up, it works 24/7 without manual clicking

What's next? In the next section, we'll learn about Required vs. Optional steps — how to make your automations flexible when some data might not always be available.


Previous: 3.6 Telegram IntegrationNext: 4.2 Required vs Optional Steps

Built with VitePress