1. Agents vs. chatbots (clear difference)
A chatbot answers questions. An autonomous agent pursues a goal across multiple steps and systems.
Example:
- Chatbot: “What is our policy for late vendor invoices?”
- Agent: Read the new invoice email → match it to a purchase order → flag a price mismatch → ask a human for approval → post the approved entry → log the outcome.
If a step fails, a good agent has a record of where it left off and of what a human needs to decide next. The record is a checkpoint, not a guess reconstructed from the chat.
2. Pillar 1 — Perception (See)
Perception means turning messy real-world inputs into clean, structured facts the rest of the system can trust.
Inputs might include:
- Emails and PDF attachments
- Rows in a warehouse or ERP table
- Webhook events from another app
- Sensor or shipment tracking messages
The perception layer’s job is not to be creative. It is to be consistent: validate fields, normalize dates and currencies, reject garbage early, and attach a confidence label when a field is not certain. The label is a score the next step can refuse. It is not a promise that the guess is right.
3. Pillar 2 — Logic (Decide)
Decision logic should be a mix of:
- Hard rules (policy, compliance, dollar limits) — fast, cheap, and predictable, and
- AI judgment only where judgment is needed (ambiguous text, unusual cases, prioritization).
A large language model is the wrong tool for a check that is already a rule. “Amount must be positive” and “approver required above $10,000” are examples. Those checks belong in code: the same answer every time, without a model call.
Good decision layers also produce an explanation a person can check. For example: “Blocked because the purchase order quantity is 10 and the invoice quantity is 12.” That sentence is what the auditor, and the person who receives the exception, actually need.
4. Pillar 3 — Action (Do)
Action is where agents create value — and risk. Every action should go through a controlled “tool” interface, similar to giving an employee access only to the buttons they need.
Healthy action design includes:
- Least privilege — the agent can update invoice status, not rewrite the entire vendor master file.
- Clear outcomes — success, bad input, rate limited, permission denied (like HTTP status codes).
- Idempotency — running the same approved action twice does not create double payments.
- Human gates — high-impact actions require explicit approval.
5. Putting the pillars together
A day-in-the-life flow for accounts payable might look like this:
- See: New invoice PDF arrives; perception extracts vendor, amount, PO number, and line items.
- Decide: Rules match the purchase order. The model notes a 2% price variance. The written policy in this example says a preferred vendor under 3% is eligible to proceed, and anything over that is an exception for the controller. The percentages are an example policy, not a Cloudadorn standard.
- Do: The agent prepares the posting package. A person still approves before money moves. The controller is pulled in when the variance is over the threshold. Under the threshold, the package is shorter. It is not an unattended payment.
Each pillar can be tested, replaced, or rolled back on its own. A single mega-prompt cannot.
6. Common mistakes to avoid
- Starting with the model, not the workflow. Pick the business process first.
- Letting the agent free-write SQL or emails with no templates. Structure outputs.
- Skipping human review for money movement. Automate preparation; gate high risk.
- No metrics. If you cannot measure cycle time and exception quality, you cannot improve.
- Boiling the ocean. One workflow done well beats five half-finished pilots.
7. How to start this quarter
A practical 90-day path:
- Weeks 1–2: Choose one workflow; interview the people who live it; map systems and pain points.
- Weeks 3–6: Build perception + rules first; add AI only where needed; wire one or two safe actions.
- Weeks 7–10: Run in shadow mode (agent recommends, humans still click).
- Weeks 11–12: Turn on only the actions a written policy already allows without a person. Keep approval on money movement and other high-risk steps. Show leadership what the shadow period measured, including where the agent was wrong.
On this path, Cloudadorn’s Forward Deployed Engineers can sit with your team. The engagement includes a handover, so the system is not meant to stay a black box after launch.