A team of planning agents, on duty 24/7
Sense → reason → decide → act → learn. Agents handle the routine exceptions that eat your planners’ day and bring humans in only for decisions that matter.
Layer 3 · AI Planning Agents
Planning agents that work 24/7 — and show their work
The ontology provides understanding. The optimizer provides answers. Agents combine the two into a system that runs on its own.
Convect runs a team of specialized agents. They aren’t chatbots or recommendation widgets — they sense business events, reason about impact, call the optimizer, and act within the authority you give them.
Demand Sensing Agent
Watches demand signals and triggers plan updates when they shift
Monitors orders, sell-through and external signals like weather and promotions
Scheduling Agent
Builds the best feasible schedule from live constraints
Calls the optimization engine with current capacity and material status
Disruption Response Agent
Replans around breakdowns, late materials and rush orders
Listens for execution events and triggers incremental replans
Coordination Agent
Keeps procurement, production and logistics plans aligned
Runs cross-functional constraint checks and resolves conflicts
Root-Cause Agent
Answers “why can’t this order ship on time?”
Traces the ontology backward to the binding constraint
The agent loop
AGENT LOOP
24/7
always on
1Sense
Continuously watch orders, demand signals, capacity, inventory and supplier updates
2Reason
Use the ontology to work out what a change means and everything it touches
3Decide
Call the optimization engine to generate and compare response options
4Act
Update the plan within its authority — or send a recommendation for approval
5Learn
Learn from outcomes and planner choices to improve the next decision
Live example: five agents handle a rush order
Multi-agent workflow · Rush order
What this means for you
Your planners stop watching screens for rush orders and late trucks. Agents handle the routine exceptions and bring people in only for decisions that matter — planners go from firefighting to setting strategy.
Ask your plan anything
Answers you can check, not just answers that sound right
Ask in plain language. Agents answer by tracing the ontology and running the optimizer — and show their work.
“Can we ship unit 14 two months early?”
Checks contract milestones, shop capacity and engineering readiness along the ontology, and answers with the chain of reasoning behind it.
“Why is SO-88 going to be late?”
Traces backward to the binding constraint — e.g. “Line 3 is full this week and the only alternate line lacks Mold M7.”
“What happens if Supplier S slips by a week?”
Runs a what-if through the optimizer and lists the orders, customers and margin at risk, with mitigation options.
Governed autonomy
Autonomous — and always under control
In an enterprise, every agent action has to sit inside rules and on the record.
Configurable autonomy
Set thresholds by cost impact, orders affected or customer tier. Inside them, agents act; outside them, they ask.
Every recommendation explained
Each proposal shows the binding constraints, cost and service impact, and how it compares with the options it rejected.
Full audit trail
Events sensed, reasoning path, optimization runs and final decisions are all logged — replayable and auditable.
Humans stay in the loop
Planners approve or edit significant decisions with one click, and agents learn their preferences over time.
Frequently asked questions
How are Convect agents different from a chatbot or copilot?+
A chatbot answers questions. Convect agents sense business events, reason about their impact using the ontology, call the optimization engine to produce feasible plans, and then act or ask for approval. They participate in decisions — they’re not just a conversational interface.
Will agents change our plan without asking?+
Only within the limits you set. Every type of action has configurable autonomy and approval thresholds — by cost, orders affected or customer tier. Most customers start in recommend-only mode. Every automated action is logged and can be rolled back.
Which language models do you use, and do they make the numbers?+
Language models handle understanding requests and explaining results. The numbers — quantities, dates, assignments — always come from the optimization engine working on your ontology, so they satisfy your constraints by construction.
Do we need real-time data?+
Real-time data makes agents faster, but it isn’t a prerequisite. Many customers start with scheduled syncs (hourly or per shift) and add MES, WMS and other execution events over time.
Let agents take the daily firefighting
Book a demo to see rush orders and equipment breakdowns handled automatically on a scenario from your industry.