D2DDiagrams-to-Dynamics, made usable

System dynamics modeling, simplified for domain experts.

A minimalistic, web-based editor that lets you visualize, configure and compare scenarios from auto-generated models — without becoming a trained modeler.

CLD → ODEPipeline
In-browserReact frontend
N scenariosSide-by-side
Uploaded Model/Baselinedynamix.app
Customer BacklogOrder ArrivalFinished GoodsRaw MaterialShipment RateRevenue RateCash BalanceWorkforceCapacity LimitSTOCK · 80% CONFinitial · 5,000 units
high medium low derivedconfidence · 74%
For complex systems thinkersDesigned for researchers, epidemiologists, and policy makers tackling complex societal problems.
epidemiologypublic policysupply chainecologyeconomics
Why Dynamix

Three things legacy modelers never gave you.

Stella and Vensim were built for trained modelers. Dynamix is built for the people who actually have the questions.

StockFlowAuxiliary
01 / Approach

No formal modeling required.

Leave visually dense tools like Stella and Vensim behind. The interface is deliberately targeted at domain experts who have a causal loop diagram and want answers — without becoming trained modelers to get them.

0k3k5k8k10k04590135180
02 / Compare

Side-by-side scenario comparison.

Real domain questions are scenario-based — what happens if we intervene here versus there? Configure many scenarios, run them simultaneously, and compare their time-series outputs to make causal differences visible.

infection_rate92%policy_lag68%budget_uptake34%
03 / Uncertainty

Confidently handle uncertainty.

Auto-generated models aren't perfect black boxes. The interface surfaces confidence scores, expected ranges and repair reports so you can easily identify and correct suspect variables or equations.

How it works

From a diagram to a simulated answer, in three moves.

A workflow that mirrors how domain experts already think — inspect the structure, sanity-check the values, run the scenarios.

STEP 01 — IMPORT

Import & inspect.

Upload your Diagrams-to-Dynamics (D2D) generated model to instantly view a simplified, interactive stock-and-flow visualization. Stocks, flows and auxiliaries are laid out cleanly — not buried in dense node-graph chrome.

model.d2d
Customer BacklogOrder ArrivalFinished GoodsRaw MaterialShipment RateRevenue RateCash BalanceWorkforceCapacity Limit
STEP 02 — CONFIGURE

Configure & correct.

Adjust initial values and parameters using built-in context, suggested defaults, and confidence metadata. Flagged variables are surfaced immediately; you decide whether to accept the LLM's best guess or override it.

parameters.scenario
PARAMETERVALUEDEFAULTC
production_rate_param0.700.7
hire_rate_param0.050.05
Fshipment_rate_param0.100.1
cost_per_unit_param10001000
desired_finished_goods…0.200.2
rate_param0.100.1
STOCK · MEDIUM
Finished Goods Inventory
MODEL CONFIDENCE
0%80%
STEP 03 — COMPARE

Compare & analyze.

Build alternative scenarios, run the simulations, and compare the outputs to see exactly how your interventions propagate through the system over time. Every variable gets a chart; every chart can be overlaid.

compare.run
SCENARIOS
Baseline
High demand+30% orders
Supply shockraw -50%
Baseline
High demand
Supply shock
t = 0 → 180 days
0k3k5k8k10k04590135180
Features

Built for comprehension, not modeling certification.

Contextual parameter configuration

Understand what each value represents and why the suggested default is reasonable — before making adjustments.

F.01

Built-in repair reports

View internal constraint-based repair notes generated by the pipeline to understand how your model was structured.

F.02

Confidence scoring

Instantly spot uncertain parameters with visual indicators for confidence scores and expected ranges.

F.03

Comparative plotting

View the outputs of two or more scenarios side-by-side on interactive time-series charts.

F.04

Minimalistic visual language

A stock-and-flow structure designed for comprehension by newcomers — without the visual density of legacy tools.

F.05

Browser-based execution

Run ordinary differential equation simulations entirely within a responsive React frontend. No desktop installations.

F.06
Ready when you are

Ready to answer "what if"
for your complex system?

Start exploring your causal loop diagrams dynamically — from structure to scenario in a single workspace.

Quick start

Getting started: from diagram to simulation.

Learn how to read your model, adjust parameters, and run your first scenario comparison.

EST · 4 steps · ~8 minutes
STEP 01 OF 04

Reading the stock-and-flow canvas

Your imported Causal Loop Diagram (CLD) is now a dynamic model. Click any node to see whether it is classified as a stock (an accumulation) or a flow (a rate of change). Causal links appear as dashed connectors; flow valves indicate where stocks accumulate or drain.

canvas.dynamix
Customer BacklogOrder ArrivalFinished GoodsRaw MaterialShipment RateRevenue RateCash BalanceWorkforceCapacity LimitSTOCK · 80% CONFinitial · 5,000 units
TIPHover any node to preview its connections; press ⌫ (delete) to flag a suspect variable.