Objectives and KPIs
What the company wants to achieve and how it’s measured — as a number, a % or a $ amount.
Actions
What the humans did (decisions, changes) — and the impact of it, assessed by the AI.
Results
How the KPIs evolve over time — the company’s results meeting, always up to date.
Learnings
“For problem X, doing Y fixed it / made it worse / changed nothing” — the memory of what we’ve already tried.
The Cortex cycle, in one sentence
You set Objectives (with KPIs) → cascade them into each agent’s objectives → record the human Actions → the AI measures the KPIs and assesses the impact of the actions → this becomes Results and Learnings → next time around, the agents already know what works.
📸 Screenshot to insert [/en/cortex/overview]: sidebar open on the 'Cortex - Business Management' group showing Objectives, Actions, Meetings, Results, Learnings and Knowledge.
Why this matters (the benefits)
Everyone (and every AI) pulls in the same direction
Everyone (and every AI) pulls in the same direction
When the objectives live in Cortex, they’re automatically injected into the agents’ context. The agent stops answering “in a vacuum” and starts acting knowing what the company’s and its area’s goal is. It’s the alignment that usually gets lost between leadership and day-to-day work.
Evidence-based decisions, not guesswork
Evidence-based decisions, not guesswork
A KPI isn’t a number someone types in once and forgets. The AI measures it periodically using the connected data (spreadsheets, databases, integrations) and shows current value × target × trend. You debate facts, not opinions.
The impact of what people do becomes visible
The impact of what people do becomes visible
Every change a human makes (raising a price, switching a supplier, changing a creative) can be recorded as an Action. The AI investigates the data to see whether it improved, worsened or didn’t change the KPI. You finally know what was worth it.
The company stops repeating the same mistakes
The company stops repeating the same mistakes
Every assessed action becomes a Learning in the format problem → action → outcome. The next time the same problem shows up, Cortex (and the agents) remember what’s already been tried — and what worked.
The cascade — from the top all the way down
Cortex works at two levels that fit together:1
Company objective
The big objective, with “top-level” KPIs (e.g. “Grow revenue 30%”, KPI: Monthly revenue).
2
Agent objective (derives from the company's)
Each strategic agent has objectives that derive from a company objective. The agent’s KPI is tied to a company KPI — so progress at the edge adds up at the top. That’s the cascade: moving the number below moves the number above.
This link (“Derives from” + “Tie to a company KPI”) is what makes the objective tree make sense. See the step by step in Objectives.
Ask Cortex
On every management screen there’s an Ask Cortex button. Cortex is a native management agent (every customer has one) that already sees your objectives, KPIs, actions and results. Ask in plain language — “how are we doing this quarter?”, “which action had the most impact?” — and it answers like a chief of staff, calling on the areas’ strategic agents when the question is specific.Who sees what (permissions and areas)
Cortex respects each person’s permissions:- Owner/Admin see the whole company.
- Members see the objectives/actions/results of the areas they belong to; private content of an area that isn’t yours doesn’t show up — not for you, and not in Cortex’s answers.
- Configuring/running objectives and analyses requires permission (owner/admin for the company; whoever “configures” the agent, for that agent).
Where to start
1
Set 1 to 5 company objectives
Start small. In Objectives, create the top-level objectives with measurable KPIs.
2
Cascade down to the agents
Give each strategic agent an objective that derives from the company’s, with a tied KPI.
3
Record and assess the actions
In Actions, log what was done and click Analyze impact.