⚡📘 Automatic Resource Management with AI Studio: The Complete Implementation Guide

Hi,
I’m Ethan DeWaal, Strategy & Ops Lead for the AI Studio team at Asana. Today we’re digging into one of the most powerful ways to use AI Studio: automatic resource management.

From simple rule-based routing to intelligent capacity-aware assignment—choose the level that fits your needs.

The Core Concept

AI Studio can now read Universal Workload data and team custom fields to make intelligent routing decisions. This means your tasks can automatically assign themselves to the right person based on:

  • Real-time capacity (who actually has bandwidth)
  • Skills and expertise (who knows how to do it)
  • Objective criteria (department, seniority, timezone)

Four Implementation Levels

Level 1: Hard-Coded Routing (5 min setup)

  • What you need: Nothing—just write the rules
  • What you get: Deterministic category-based routing
  • Best for: Small teams, clear categories, predictable work types
  • How it works: Write explicit “if-then” rules in your instructions

Level 2: Skills-Based Routing (15 min setup)

  • What you need: A team with custom fields
  • What you get: Dynamic routing based on skills/attributes
  • Best for: Teams with varied expertise, skill-based work
  • How it works: @mention a team, AI reads custom fields and routes

Level 3: Capacity-Only Routing (20 min setup)

  • What you need: Universal Workload view
  • What you get: Workload-balanced assignment
  • Best for: Similar skills, preventing burnout, even distribution
  • How it works: Paste workload URL, AI assigns to whoever has bandwidth

Level 4: Full Intelligence (30 min setup)

  • What you need: Universal Workload + Team + Custom Fields
  • What you get: Capacity AND skills-based routing
  • Best for: Complex teams, high volume, nuanced decisions
  • How it works: Combined data for smartest possible routing

Level 1: Hard-Coded Routing (The Simplest Start)

Sometimes you don’t need AI to be smart—you just need it to follow rules. This requires ZERO data access.

Basic Template

Category-Based Template

Pros of Level 1:

  • Zero setup beyond writing rules
  • Completely predictable
  • No data permissions needed
  • Works immediately

Cons of Level 1:

  • No capacity awareness (might overload people)
  • Can’t adapt to team changes
  • Rigid—needs manual updates

Example: Our CRM team started here. Every Salesforce request went to Beverly, every web task to Amy. It worked, but Beverly got overwhelmed during busy periods.

Level 2: Skills-Based Routing (Adding Intelligence)

When you need dynamic routing based on team attributes but don’t need capacity data.

Setup:

  1. Create/update your team
  2. Add custom fields (Skills, Department, etc.)
  3. @mention the team in your rule

Template with Skills Matching

What AI sees: Up to 50 team members with all their custom field data

Example: Now our rules could say “find someone with Salesforce skills” instead of “assign to Beverly”—helpful when we hired another Salesforce expert.

Level 3: Capacity-Only Routing (Workload Balance)

When everyone can do the work, but you need to balance the load.

Setup:

  1. Create Universal Workload view
  2. Add relevant people
  3. Paste URL in instructions

Pure Capacity Template

What AI sees: Real-time capacity for up to 200 people

Example: Our content team uses this—everyone can write, so we just need even distribution.

Level 4: Full Intelligence (The Complete Picture)

Combining everything for sophisticated routing decisions.

Setup:

  1. Create Universal Workload view with your people
  2. Create team with same people
  3. Add custom fields to team
  4. Reference workload URL in instructions (by referencing the workload, it gives access to the public team level custom fields for the user, across all teams they are apart of).

Advanced Multi-Factor Template

What AI sees:

  • Capacity data for all people in workload
  • Custom fields from their team memberships
  • Full context for intelligent decisions

Example: This is where our CRM team is now. Beverly gets Salesforce tasks ONLY when she has capacity. If she’s swamped, it goes to our backup person with Salesforce skills.

How to Choose Your Level

If you have… Start with… Because…
<5 people, clear roles Level 1 (Hard-coded) Simple, predictable, no setup
Varied skills, stable workload Level 2 (Skills) Matches expertise without complexity
Similar skills, varying workload Level 3 (Capacity) Prevents burnout, ensures balance
Complex needs, high volume Level 4 (Full) Maximum intelligence, best outcomes

Migration Path: Level Up When Ready

Week 1: Start with Level 1

  • Hard-code your basic rules
  • See immediate automation value
  • Learn what patterns emerge

Week 2-3: Move to Level 2 or 3

  • Add either skills OR capacity
  • Test which matters more for your team
  • Refine your rules

Month 2: Graduate to Level 4

  • Combine both data sources
  • Add sophisticated logic
  • Handle edge cases

The Technical Details

Data Limits by Level

Level Data Access Limits
1 None No limits—just your rules
2 Team fields 50 team members max
3 Workload 200 people + their team fields

Critical Constraints

  • Cannot @mention workload views (must paste URL)
  • Date logic: Tasks with dates use those dates; no dates = 4-week window

Start Where You Are

Don’t overthink it. If you’re manually assigning more than 5 tasks a day, even Level 1 will transform your workflow.

Pick your level. Copy the matching template. Customize for your team. Deploy today.

The right work will start finding the right person—automatically.



:inbox_tray: Get notified of new posts :backhand_index_pointing_down:

Want to be automatically notified when new AI Studio Tips go live?
:bell: Make sure to follow AI Studio Tips & Workflows! :bell:



9 Likes

This is awesome! I didn’t even realize teams supported custom fields for people…gotta brush up on all these great new features

3 Likes

Do we have any details on credit usage for these rules to run? I’d love to build a case in our company for adopting the AI pro bundle - this could be really helpful.

Hi @Simon_Koekkoek, The total number of credits consumed by each workflow is calculated by multiplying the content size by the model multiplier (see model pricing here). You could check the rule history and see how much each rule execution costs.

I also sent you a direct message with some extra help our team can offer!

That link was very useful thank you!

@Ethan_DeWaal does it only see “current” capacity right now, or can it designate a person for a task to do next month?

Great post by the way, thanks :slight_smile: