SIMWELL COMPASS

Your models.
Your team's hands.

A model that lives on one analyst's desktop answers a question once. Compass is where it becomes a decision system: a tool your planners open, run against this week's numbers, and trust enough to act on.

It runs models your team builds, models SimWell builds, and models you bought from someone else. If a model exposes an API or runs as an executable, Compass can run it.

The Compass scenario workspace: a sortation model with five saved scenarios, version control, and run controls, all in the browser.

01 / THE OPPORTUNITY

Models belong in the hands of decision-makers

The models that could guide a decision are almost never at the fingertips of the people making it. They stay in the hands of specialists, locked inside desktop tools and one-off files.

A planner who needs an answer before Thursday's commit meeting does not open the model. They ask someone, and they wait. Decisions wait on whoever can run it, answers come back slowly, and most of the value never reaches the people who need it.

What changes when the model is within reach

  • Models open in a browser, with no local software for anyone who needs to use one
  • Scenarios are built and compared by the planners who own the decision
  • Every result traces back to a specific run, model version, and set of inputs
  • One audit trail and one set of access rules across every model you run
  • The specialist still builds the model. Everyone else can now run it.

02 / THE PLATFORM

One platform, every decision horizon

Deploy models against every decision your business makes, from the operational calls in front of you right now to the commitments that shape the next decade.

NOW

Operational

React to what is happening now. Run the model against current conditions and decide before the shift turns over.

THIS WEEK

Tactical

Deliver this week's plan. Test the sequence, the staffing, and the routing against the volume actually coming.

NEXT QUARTER

Planning

Prepare for next quarter's demand. Compare options against a forecast rather than against last quarter's average.

THE NEXT DECADE

Strategic

Shape the years ahead. Put capital and capacity decisions through the model before the commitment is irreversible.

Governed by default

Version control, usage tracking, and role-based access across every model on the platform, with one audit trail rather than one per tool.

Any engine, any language

Optimization, simulation, and forecasting models run side by side. If a model exposes an API or runs as a standalone executable, Compass can run it.

Ready for live data

Built to connect to live feeds, so runs reflect current conditions rather than last quarter's extract.

03 / WHY WE BUILT COMPASS

A need we felt, and saw in our customers

Our need

AS CONSULTANTS

We deliver simulation and optimization across many projects, teams, and regions. We wanted to work from a consistent foundation rather than starting each engagement from scratch.

A shared platform means solutions are faster to stand up, easier to support, and what we learn on one project carries into the next.

Our customers' need

AS OPERATORS

Our customers face the same challenges we do. One difference matters more to them than consistency or speed.

Getting these tools out of specialists' hands and close to the people making the decisions, on top of the consistency, speed, and reuse a shared platform brings.

SO WE BUILT COMPASS

One platform to put models to work

For our teams and our customers alike. We run our own delivery on it.

Start a conversation

04 / COMPASS IN ACTION

Create, run, and compare scenarios in the browser

Every model becomes a governed, versioned tool your team runs on demand. No desktop software, and no queue behind one analyst.

Compass scenario list for a sortation model, showing five scenarios with owner, creation date, run status, and per-scenario actions.
One sortation model, five scenarios, each owned by a named user and pinned to a model version. Selecting two or more compares their results side by side.

And models can feed each other

Forecasting
demand
Optimization
a plan
Simulation

Forecast demand, build an optimized plan around that forecast, then test the plan against real operating conditions. Each model's output becomes the next model's input, so the chain runs as one job rather than three handoffs.

05 / THE AI ASSISTANT

A team of agents, with you in command

Compass puts specialized agents to work across your models. Each one takes a job, from designing scenarios to running experiments to explaining what drove a result, so your team gets from question to answer in an afternoon.

SCENARIO AGENT

Describe what you want to test in plain language. It designs and builds the scenarios for you.

EXPERIMENT AGENT

Sets up a design of experiments to explore the space efficiently, instead of running every combination by hand.

INVESTIGATION AGENT

Interrogates your scenarios and results to explain why an outcome occurred and what drove it.

INSIGHT AGENT

Turns runs into plain-language explanations and draft reports your whole team can act on.

DATA AGENT

Prepares and edits the underlying input data, so models are ready to run without manual wrangling.

HUMAN IN THE LOOP

An assistant, not an autopilot. Every run is authorized by a named person and every recommendation is yours to accept. The agents do the legwork. The decisions stay with you.

GROUNDED IN YOUR MODELS

Every answer comes from your own model, not from a language model's general knowledge. The agents design an experiment, run your validated model, and report what came back. Every recommendation traces to a specific run, model version, and set of inputs, so anyone can reproduce it and check it.

06 / THE ASSISTANT AT WORK

Ask a question, get an answer you can act on

A real run on an automated storage and retrieval system. Fifteen scenarios, two levers, and a conclusion that changed what the team was about to buy.

A Compass AI answer: fifteen scenarios designed, sensitivity read from the runs, erector speed identified as the main driver at plus 2.8 percent, elevator delay secondary at minus 2.2 percent.
The conclusion is the useful part: a 200% erector speed increase lifts output only 2.8%, so the system is already well balanced and a large upgrade would hit diminishing returns.

07 / FOR TECHNICAL TEAMS

Built to deploy, govern, and scale your models

What your architects, security reviewers, and platform team will want settled before this goes anywhere near a procurement conversation.

INTEGRATION

Enterprise integration

Build pipelines and connections into the systems you already run on, so Compass fits your stack instead of asking your stack to accommodate it.

EXTENSIBILITY

Custom scripts and logic

Run scripts before, after, or around model execution to add logic, transform inputs and outputs, and shape the workflow to your process.

SECURITY

Security and access control

Role-based access decides which models a person can open and run. A full audit trail records every run against a named person, so every result is attributable.

GOVERNANCE

Governance and collaboration

Built-in BI dashboards for live results, with version control and auditability across models, inputs, and outputs, so anyone can reproduce a number and check it.

Hosted by SimWell

MANAGED FOR YOU

We run and manage Compass so your team can start without standing up infrastructure, and every upgrade is handled.

  • The fastest way to get started
  • No infrastructure to manage
  • Always on the latest release

In your own environment

DEPLOYED IN YOUR CLOUD

Compass runs inside your infrastructure, where your models, inputs, and results never leave it. Point it at AI models you host yourself, so nothing sensitive crosses your boundary.

  • Your data stays in your environment
  • Your identity, networking, and data residency
  • Your choice of AI models, kept in-house

08 / COMMON QUESTIONS

Questions buyers ask before they engage

Do we have to rebuild our models to run them on Compass?+
No. Compass runs models in whatever engine or language they were built in. The requirement is that a model exposes an API or can be run as a standalone executable.
Does SimWell have to build the models?+
No. Compass runs models your team develops in-house, models SimWell builds for you, and models built by anyone else. Many customers run a mix.
How does Compass relate to the simulation software we already license?+
Compass does not replace it. Your modelers keep building in the tool they know. Compass is where the finished model becomes something the rest of the organization can run.
Can our planners build scenarios without an analyst?+
Yes. Scenarios are created and run in the browser, and the scenario agent will build a set from a plain-language description. Every run is authorized by a named person and traces back to a specific model version.

WHO THIS IS FOR

A model your team should be running, that only one person can run.

Compass is for operations groups who already have models they trust and cannot get them in front of the people making the call. If your planners wait on an analyst, if a decision goes stale before the run comes back, or if nobody can say which version produced last quarter's number, that is the problem this solves.

See it on your own models

Bring one model and one decision it should be answering every cycle. We will show you what it takes to stand it up in Compass, and if your model is not ready for that yet, you will hear it in the first call.

Start a conversation