A workflow is a fixed sequence of steps. An agent decides what to do at each step. Both have their place.
If the steps are always the same — use a workflow. If the steps depend on what the agent finds — use an agent.
Three trade-offs
Workflows are predictable — agents are flexible
Workflows are cheaper — agents use more tokens
Workflows are easier to debug — agents are harder
The big idea: Pick the right tool. Workflows for routine. Agents for novel.
End-of-lesson check
15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-builders-agentic-agent-vs-workflow
What is the main difference between an agent and a workflow?
Agents decide what to do at each step based on the situation
Agents follow a fixed sequence of steps exactly as programmed
Workflows can change their approach based on what they find
Workflows are more flexible than agents
A company needs a system that automatically pulls sales data every Monday and emails a summary to the team. What type of system should they use?
A workflow, because it can investigate why sales changed
A workflow, because the steps are the same every week
An agent, because it will adapt if sales numbers look unusual
An agent, because it can decide what to do based on the sales data
A system needs to investigate why website traffic dropped on a particular day, figure out possible causes, and suggest fixes. What type of system should they use?
A workflow, because it follows a predictable sequence
An agent, because it must decide what to do based on what it finds
An agent, because it sends a report on a schedule
A workflow, because the steps are the same every time
Which statement describes a key advantage of workflows over agents?
Workflows can change their plan based on new information
Workflows can handle any situation that comes up
Workflows are more predictable because they follow the same steps
Workflows use more tokens than agents
Which statement describes a key advantage of agents over workflows?
Agents cost less money to run
Agents are more flexible and can adapt their approach
Agents follow the same steps every time
Agents are easier to debug when something goes wrong
Why are workflows generally easier to debug than agents?
Workflows always produce the same output for the same input
Workflows can adapt and try different approaches
Workflows use fewer tokens so errors are smaller
Workflows are less predictable
What does 'flexibility' mean when describing an agent?
Following a predetermined sequence of steps exactly
Costing less money to operate
Being able to change approach based on what it discovers
Sending reports on a regular schedule
What does 'observability' refer to in the context of workflows?
The ability to track and monitor each step in the process
The flexibility of the workflow
The ability to see what the agent is thinking internally
The cost of running the workflow
A system sends a daily weather summary to all users at 6 AM every morning. What type of system is this?
An agent, because it processes user data
An agent, because it decides what weather to include
A workflow, because it investigates weather patterns
A workflow, because the same steps happen on a schedule
An AI system analyzes a customer complaint, searches for similar issues, determines the root cause, and proposes a solution. What type of system is this?
A workflow, because it runs on a schedule
An agent, because it sends a response at a fixed time
A workflow, because it follows clear steps
An agent, because it decides what to investigate based on the complaint
If you need to create an audit trail showing exactly what happened at each stage of a process, which would be easier to implement?
An agent, because it adapts to the situation
A workflow, because it uses fewer tokens
A workflow, because the sequence is fixed and each step is known
An agent, because it makes smart decisions
What does it mean when we say a workflow is 'predictable'?
It follows the same steps in the same order every time
It can change its approach based on new information
It handles unexpected situations well
It costs less than an agent
Why might an agent produce different results for seemingly similar situations?
Agents are limited to a fixed sequence of operations
Agents always follow the exact same steps
Agents make decisions based on the specific details of each situation
Agents cannot access external data
Which tool is best suited for routine, repetitive tasks that never change?
A workflow, because it uses more tokens
An agent, because it can adapt to each execution
An agent, because it is more flexible
A workflow, because the steps are the same every time
Which tool is best suited for situations where you don't know in advance exactly what steps will be needed?
A workflow, because it follows a predictable sequence
An agent, because it sends reports on a schedule
A workflow, because the steps are always the same
An agent, because it can figure out what to do based on the situation