Show HN: Simulate what AI agents do to an engineering org (no signup)
Meridian is a free simulator that models the impact of AI agents in a 35-person SaaS engineering org. Adjust assumptions, see real-time changes in cost, capacity, and coordination. Share scenarios as links or export as OrgSpec.
Free tool — no signup
What happens to your org when the agents arrive?
This is Meridian: a realistic 35-person SaaS product and engineering organisation. Embed coding agents in squads, replace tier-1 support, flatten a management layer, merge teams — and watch cost, capacity, and coordination change live. Then share your scenario as a link, or export it as OrgSpec.
Built on a simplified version of the model Orgonaut uses. Every assumption is visible and editable below.
How the model works
Opinionated on purpose, honest about limits
This simulator is not a forecast. It is a structural argument: adding AI agents to an organisation changes cost and capacity, but supervision, team size, and management span decide whether that capacity is real. The rules below are simplifications — deliberately visible ones, so you can challenge them rather than trust a black box.
When you want the same discipline against your real organisation — real people, real salaries, real teams, with history, approvals, and audit — that is what Orgonaut does.
Agents are not free capacity
Each senior or lead can effectively supervise a limited number of agents. Agents beyond supervision capacity contribute at 40% effectiveness. Add agents without seniors and the model tells you.
Big teams pay a coordination tax
Every member beyond five discounts team capacity by a configurable percentage. Merging squads is cheaper on paper — the model shows what it costs in throughput.
Every assumption is a slider
Agent cost, agent throughput, supervision ratios, and the coordination tax are all editable. Disagree with a number? Change it and watch the whole scenario recompute.
Deltas, not vibes
Every edit is compared against the Live baseline: cost, capacity, headcount, and cost per delivery point. The change log reads like a reorg proposal because that is what it is.
Take it further
From a shared toy to a decision you can defend
The simulator makes the argument tangible. The real version of this conversation happens against your own organisation.
For CTOs and VPs of Engineering
Share a scenario link with your leadership team instead of a hand-drawn diagram. Then model the real thing: import your org, clone it into a scenario, and compare deltas with real numbers.
For AI consultants
Run this live in workshops. Adjust the assumptions with the client in the room, agree on the numbers, and use the exported OrgSpec document as the starting point for a real engagement.
For the sceptics
Crank the agent count with no seniors, or merge everything into one mega-team. The model pushes back — because structure still matters in the agentic age. That is the whole point.
Model your real organisation Run this with clients
Design your organisation for the agentic age
Import your real org, model scenarios with people and agents together, and compare deltas before you commit anything.
Get started free Talk to Orgonaut