Agentic AI Chapter 2 of 9

An agent carries work through to an action, under your rules.

An assistant answers the question you ask and leaves the next step with you. An agent takes an objective, works through the steps, acts in your systems and records what it did, inside rules you set.

MONDAY 21 SEPTEMBER 2026 · L. TAN’S EXCEPTION LIST · SHOWCASE ESTATEAGENT · CARRIES THE WORK THROUGH TO THE ACTIONthe whole path: 38 minutesASSISTANT · STOPS AT THE ANSWERThe list1,240 exception messages;212 touch an open purchaseorder07:40 · the list opensThe answer38 lines can still becancelled this week, worth$412,0004.2 sThe draft29 drafts under $25,000 toD. Okafor; 9 to him asbuyer, for his judgement6 minThe actioneach approved cancellationwritten to the ERP andacknowledged22 min to approve · 9 min to write backUNDER YOUR RULES · CHECKED BEFORE EVERY ACTIONpolicy version 3value threshold $25,000daily budget 60confidence floor 0.80approver D. Okaforevery write acknowledged
An assistant stops at the answer; an agent carries the same Monday through to the confirmed action, and your rules decide what it may do on the way. Showcase estate.
38 min
from L. Tan’s question to confirmed cancellations once an agent may act, against 2.5 days by hand
showcase estate
8
parts make up every Juno agent, from its objective to its tests
product
130
vendors, of the thousands claiming agentic AI, judged to have genuine agentic capability (Gartner, 25 June 2025)
dated source
1
Juno agent in production today: the PO cancellation agent, at a high-tech manufacturer
production
Three shapes

An assistant answers, an agent acts, an agentic system shares one set of rules

One platform carries all three: ChatLima answers, the one built agent acts, and one rule store and one audit trail would govern every agent built after it.

THREE THINGS PEOPLE CALL AI, EACH BUILT ON THE ONE BEFORE01Assistantanswers the question you ask; the next step stayswith you02Agentcarries work to an action, toward an objective,inside limits03Agentic systemmany agents, one set of rules, one audit trail,one level each
Each shape keeps the one before it: the agent acts from the same understanding the assistant answers from, and the system would govern every agent built from the blueprint with one set of rules. Product.

On the Monday: the assistant

L. Tan asks ChatLima which open lines she can still cancel this week and what they are worth: 38 lines, $412,000, in 4.2 seconds, with the query on show. The next step stays with her.

On the Monday: the agent

The PO cancellation agent applies the policy to the same 38 lines and drafts 29 cancellations worth $187,300 for D. Okafor to approve; the 9 lines above $25,000 go to D. Okafor as buyer, for his judgement, with the reason for each.

Beyond the Monday: the agentic system

Ten task agent roles are designed on the blueprint, PO cancellation among them; one runs. Built from the same parts, they would share one rule store, one audit trail and one level setting each, so the tenth would be governed the way the first is.

The difference sits in what comes out. An assistant produces something for a person to use: an answer, a summary, a draft. An agent produces an action in a system of record, or a draft of one for a named person to approve, and it keeps the facts it acted on. The third shape is a design: as roles are built, it would hold their objectives, rules and limits in one place, so a manager would read what ran and who approved it in one report. Today that report counts one agent.

Showcase estate for the Monday’s values; the mechanics are the product’s.

One agent

Eight parts make up the PO cancellation agent, and every agent after it

This is the agent the whole story follows: the next chapters set its level, show its settings and build it in a day.

PO cancellation agentProcure-to-Pay · version 3Objectivecancel purchase order lines that standabove the need, inside their cancellationwindowData scopepurchase order lines, schedule lines,demand, stock, supplier class · plants1210 and 1220Rulesthe policy sentence and the planning tableit compiles to · version 3Trigger06:00 each working day, after the planningrunActionsdraft a cancellation · send the change tothe ERP · open a query to the buyerLimitsvalue threshold $25,000 · daily actionbudget 60 · confidence floor 0.80Escalationabove the threshold, to the buyer with thereason · a failed write-back, to the ownerTestsrecorded cases with the outcome peoplechose · last quarter: 184 lines, $1.9M
One agent, eight parts, with the showcase estate’s values filled in; the same eight parts describe every Juno agent, so a reviewer who has read one knows where to look in the next. Showcase values; the agent itself runs at a high-tech manufacturer.
The rules, in plain language

“Cancel open purchase order lines where projected excess is above eight weeks of demand, the line is inside its cancellation window, the line value is under $25,000 and the supplier class is standard. For lines of $25,000 and above, draft the cancellation and ask the buyer.”

L. Tan wrote it on the Friday; Juno compiled it into a planning table, and version 3 ran on the Monday. Showcase estate.

Objective. One sentence says what the agent works towards, and every run’s outcome is checked against it. Here: cancel purchase order lines that stand above the need, inside their cancellation window. The planner, L. Tan, owns it.

Data scope. What the agent may read, under the access rights it carries: purchase order lines, schedule lines, demand, stock and supplier class at plants 1210 and 1220, read in place from SAP S/4HANA and Oracle Fusion. The data lead sets it.

Rules. The policy in plain language and the planning table it compiles to, held as a version: version 1 the first draft, version 2 after L. Tan’s edit, version 3 the one that ran on the Monday. A change means a new version, simulated first.

Trigger. A schedule, an interval, an event in a source, or a person asking for a run. This agent runs at 06:00 each working day, after the planning run, so its drafts are waiting when L. Tan opens her list at 07:40.

Actions. A short list of what the agent may do, drawn from a fixed vocabulary: draft a cancellation, send the change to the ERP, open a query to the buyer. Everything outside the list stays outside its reach.

Limits. Checks that run before every action: a value threshold of $25,000, a daily action budget of 60 and a confidence floor of 0.80. On the Monday the threshold split the 38 lines into 29 under it and 9 above.

Escalation. The named person or desk that receives what the agent hands over, with the reason. Above the threshold, a line goes to the buyer; a failed write-back goes to the owner. The 9 lines needing judgement reached D. Okafor this way.

Tests. Recorded cases with the outcome people chose, used to release the agent and later to raise its level. Before release the policy ran over last quarter: 184 lines matched, worth $1.9M, 31 routed to the buyer, read before J. Ridley released version 3.

Insight to action

Four kinds of analysis lead to an action, and the same four steps sit under each

The same four steps serve a measure you only watch and a decision you let an agent make.

FOUR KINDS OF ANALYSIS, ONE METHOD UNDERNEATH01Descriptivewhat happened, and where it standstoday02Pre-emptivewhat is about to pass a limit youset03Prescriptivewhat to do about it, with thebasis and a confidence04Directivethe agreed action, carried outinside the limitsmanager reports count the corrections agents made and the ones people made
Each kind keeps what the one before it does and adds a step. On L. Tan’s Monday the answer is descriptive, the Thursday warning is pre-emptive, the 29 drafts are prescriptive and the confirmed write-back is directive. Method; the Monday’s values are showcase estate.
State the rule to watch, in plain languageA person writes what good looks like: forecast error stays at or under 10% for A-class materials.
Worked example: forecast error

The rule says forecast error stays at or under 10% for A-class materials. On controller board 7741-X the error reaches 14%, so the planner is alerted with the last six periods beside the figure.

A forecasting agent configured from the same eight parts would test the other forecast methods on the same history and apply the best one inside its limits; above them the planner chooses, and the choice is kept for the next cycle. Illustrative values; the method is the product’s.

The four steps are the eight parts at work: the rule to watch is the rules part, the right person is the escalation route, the limits decide whether the agent resolves or engages, and the execution is an action from its list.

Five tests

Five tests tell you whether an agent is genuine

Each test has a screen behind it in the reference pages, so an analyst can verify the answer in the product.

Gartner counts about 130 vendors with genuine agentic capability among the thousands describing their products that way, and expects over 40% of agentic AI projects to be cancelled by the end of 2027 (Gartner, 25 June 2025). Five plain tests let a buyer check any vendor’s claim, this one included.

It works towards a stated objectiveThe PO cancellation agent carries its objective as its first part, and Juno checks the outcome of every run against it. Outcome checks
It plans several stepsThe Planner Agent broke L. Tan’s sentence into five rows: three checks, one action, one limit. She adopted four and edited one. Planner Agent
It acts in your systemsApproved cancellations go to the ERP through its supported interfaces, and the ERP acknowledges each one. Write-back
It improves from resultsEvery approve, edit and reject is kept with the facts of the moment; after six weeks, 203 of 212 drafts had been approved as written. The learning loop
It works under oversightA named person approves each action, the limits are checked first, and every action resolves to a rule version on the audit trail. Audit trail

Showcase estate for the figures in the five rows; the mechanics are the product’s.

The open question

Two of the eight parts depend on a decision the anatomy leaves open

Limits and escalation are two of the eight parts, and both depend on how much the agent may do alone.

SET FIRST, FOR THIS AGENTHow much this agentmay do on its ownto be setinformsdraftsexecutesself-correctsper agent · by a named personLimits: where itsreach endsvalue threshold $25,000 · dailyaction budget 60 · confidencefloor 0.80Escalation: who takesover beyond itabove the threshold, to thebuyer with the reason · afailed write-back, to the ownerDecided agent byagentby a named person, onevidence: the next chapterchapter 3
The threshold says where the agent’s reach ends; the escalation route says who takes over beyond it. Both follow from one setting, made agent by agent, which the next chapter explains. Showcase values.

The $25,000 threshold says where the agent’s own reach ends; the escalation route says who takes over beyond it. Both answer one prior question: how much this agent may do on its own. Set that differently and both parts change with it.

Chapter 1 named unclear limits as one of the four reasons earlier AI stalled. The eight parts give the limits a place to live. They leave open who sets the reach, on what evidence, and whether it can differ between a cancellation agent and a forecasting agent. That is the next chapter.

Terms on this page
AssistantSoftware that answers a question or drafts text and leaves the next step with you.
AgentSoftware that works towards an objective under your rules, and drafts or carries out an action.
Agentic systemMany agents sharing one rule store, one audit trail and a level setting for each of them.
LimitsChecks that run before every action: value thresholds, a daily action budget, a confidence floor.
Escalation routeThe named person or desk that receives what an agent hands over, with the reason.
Questions
Is agentic AI the same as the workflow automation we already run?

A workflow follows a fixed path that someone coded. An agent takes an objective and versioned rules, works out which steps fit the case in front of it, and reports what it did against that objective.

Does the language model make the decision?

The rules engine decides. The language model reads the planner’s intent at design time, explains answers and proposes rules it finds in past decisions. Every action resolves to a rule version and a named approver.

Does an agent replace the planner?

The agent takes the search, the valuation and the draft. The planner keeps the exceptions, the supplier relationships and the judgement calls, which is where the 9 lines of chapter 1 sit.

In the platform

Where Juno sits among the four capabilities

Each one builds on the one before it.

One shared foundationCatalogLineageAccess policyWrite path

See Juno answer and act on your own data

Bring two systems and one decision your planners make every week. In the first week we register the data, confirm the mapping with your team, answer your questions in ChatLima and run the first agent at the level you choose.