Roadmap

MySolvers is a discrete-event simulator built around one insight that sets it apart: you can tune arrival rates, service rates and queue capacities while the model runs and see the effect immediately — no stop, edit, re-run cycle. The roadmap below takes that foundation and turns it into a complete decision platform.

No release dates. A roadmap without dates ages well; one with dates does not.

Now
IN PROGRESS

Undo and redo

Undo and redo across every edit.

Next
NEXT

Results export

Take a full statistical summary away as CSV or JSON.

NEXT

Distribution library

Choose the distribution behind every rate and duration, not just the exponential default.

NEXT

Resource pools

Model a connection pool or thread pool as one component with a capacity, and let requests hold it across several steps.

Later
LATER

Traffic profiles

Drive a model with a load pattern that changes over time instead of a constant arrival rate.

LATER

Routing rules

Split traffic by percentage, by request type or by priority, not only round-robin.

LATER

Retries and backoff

Model the retry storms that turn a small timeout into a cascading failure.

LATER

Request types

Give entities a class and attributes, and let service demand depend on them.

LATER

Scenarios

Define parameter sets, run them all, and compare the outcomes side by side.

LATER

Sub-models and a component library

Collapse part of a model into a reusable block, and start from ready-made components.

LATER

Event trace

Step through a run event by event to see exactly why a model behaves as it does.

LATER

Cost modelling

Attach a cost to each resource and read cost next to latency in every result.

LATER

Cloud models

Save, share and version models against your account, and build one from your existing topology.

Exploring
EXPLORING

AI parameter optimisation

Search millions of configurations for the point where performance meets minimum cost.

Shipped
SHIPPED

Batch mode

Run a model at full speed for a set length of simulated time and collect results without watching it play out.

SHIPPED

Replications and confidence intervals

Run a model many times and read every result as a mean with a 95% confidence interval, not a single sample.

SHIPPED

Reproducible runs

Every batch starts from a seed; the same seed gives the same results, so two configurations can be compared on equal terms.

SHIPPED

Warm-up control

Discard the start-up transient so results describe the system in steady state, and find how much from pilot runs.

SHIPPED

Goodness-of-fit testing

Test the cycle time distribution against an exponential or normal with Kolmogorov–Smirnov.

SHIPPED

Live traffic

Watch traffic move along every connection as the model runs, and see where a full queue turns it away.

SHIPPED

Delay component

Model transport time or a fixed lead time between two stages.

SHIPPED

Model library

Open the classic queueing models, prepared and ready to run.

Want to know when something ships?

Leave your address and I will write when an item on this page is done. No release dates, no newsletter.

Occasional email when something here ships. Nothing else, and you can ask to be removed at any time.