talk to usGet a plan you can defend, for ports, grids, plants and rail. Planners describe how they decide. kint builds the model, and the solver's gap shows how far the plan can be from the best one under your rules. A person approves it, and the record keeps the name.
Today, plans with a solver's bound and gap belong to the few organisations with an operations research team. kint brings that method to planning teams: one engine for any decision with rules, a goal and data.
Planners decide daily under hard limits. Spreadsheets and experience produce a plan. Nobody can show whether a better plan exists, or explain the choice, when the board, the regulator or the auditor asks.
The experts who run the operation bring its knowledge. kint brings the mathematics and keeps the record.
Read-only exports from the ERP system, the terminal system or a spreadsheet. kint keeps every upload as a version.
Planners explain the goal and the rules in their own words. The decision agent writes them down for review.
The agent generates an optimization model from the rules. Domain experts check it before it runs.
What if demand rises, a crane fails, the budget is cut? Every variant runs on the same engine.
kint shows the plan, the gap that says how far from best it can be, and the limit that binds it. A person approves, and kint keeps the record.
Written rules are solved exactly. Behaviour nobody can write as a formula is learned from history, with its uncertainty. kint joins both into one model, and level by level when a decision is nested, like berth, crane and yard.
Each decision then runs as a workflow: data in, model solved, rules checked, plan released by a planner, every step recorded.
Load an example and look around. You see what needs the team today, the latest plan with its figures, the agent asking before it acts, and each run on record.
Loading the product screen
Ports, grids, factories, ERP planning and hospitals show the range. Any decision with rules, a goal and data fits the same engine: a new model, the same checks, the same record.
How many cranes and carriers does each shift need so every ship leaves on time, with the fewest machines?
Which redispatch measures relieve a congested line at the lowest cost, without risking supply security?
In which order do the jobs run on each machine, so changeovers stay short and due dates hold?
Which plant and which line take which orders next quarter?
Who works which shift next month, so every night has a senior nurse and wishes are met where the rules allow?
Rail, public transport, municipal utilities and more run on the same engine: all use cases or bring your decision.
A forecast tells you what might happen. kint tells you what to do, and how close the plan is to the best one the rules allow.
Regulated, public or private: accountability runs inside kint, from the first upload to the approved plan.
Research, software architecture and AI engineering in one founding team. You speak with the founders and with the forward deployed engineers who build the model with you.
more about the founders
Chief Executive OfficerProf. Dr. Pascal Laube
Chief Technology OfficerPedro Castelo Branco Lourenço
Lead AI EngineerJozef MikytaNo large IT project and no replacement of existing systems. kint sits next to what you already run.
One day with the planners. Together we pick one recurring decision and write down its goal, rules and data.
kint's forward deployed engineers (FDEs), engineers who work directly with the customer, build the model with the planners and compare its plans with today's.
The approved workflow runs whenever your systems call it. Each plan waits for release, then goes to the planning system through an API you control.
A conversation with kint's FDEs about one decision and its rules. You leave with a first model your planners can read.