The control plane between a decision and the four tools it touches.
Conductor is decision execution infrastructure for ops and engineering leads at 20–150 person companies. A directive becomes typed, owned, policy-checked work in Linear, Slack, Notion and Gmail — and one audit trail across all of them.
A decision made on Tuesday is gone by Friday.
A head of operations decides in a Tuesday meeting: cut the Q3 paid budget by 40%. The rationale lives in her head. The instruction goes into a Slack thread. The tickets never get created. By Friday, marketing is still spending, finance hasn’t sent the final number, and the replacement plan is a bullet point in someone’s notes app.
This happens constantly, and it’s expensive. Not because people are careless — because turning a decision into correctly-assigned, tracked, cross-tool work is manual, inconsistent and untraceable. Accountability quietly evaporates.
Three things an ops lead deals with on a Monday.
Work gets created in four places and governed in none.
A decision made in Slack becomes a Linear ticket, a Notion page and an email to a vendor. Each tool tracks its own object. Nobody holds the through-line: what was decided, what it caused, and who approved the part that left the building.
Agents multiplied the problem they were meant to solve.
Every tool in the stack now ships an agent that can create and act. Ops leads inherited a question none of those tools can answer: what did all of these just do, under whose policy, and who signed off?
The follow-through still depends on somebody remembering.
The thread scrolls. The ticket gets filed without the reason. Six weeks later the same decision gets relitigated because the rationale lived in one person’s head and nothing wrote it down next to the work.
Card 02 is the one that got worse in the last year. That’s what makes this a now.
Every tool stops where the decision ends.
Notion stores knowledge. Jira stores work. Slack distributes messages. Zapier moves structured data on deterministic triggers. Generic AI chat generates ungoverned text. Each is good at its job — and each stops exactly where accountable action starts.
Decision execution infrastructure is the layer between a decision being made and the work it causes being tracked, owned, and done — with a policy check in front of every action and one record of who approved it, across every tool it touches.
This is a structural gap, not a feature gap. Nobody owns it because the underlying capability — reliable structured output from messy human intent — only recently became real, and because the vendors best placed to build it are the ones structurally forbidden from being neutral.
The timing is structural, not narrative.
Structured output crossed a usability threshold
Unstructured intent can now become typed, executable structure reliably enough to trust with real actions. Every incumbent category was designed before that was true.
It is also commodity, and we say so. This is the premise, not the moat.
Every vendor shipped an agent. Nobody shipped a referee.
In the last year Linear, Notion and Slack all shipped agents that create and act inside their own product. Not one of them can govern an action in a competitor’s tool, and none of them will — being neutral across the stack is the one thing their business model forbids.
The scarce layer moved from execution to oversight, and it moved in our direction.
One directive in. Owned work out.
The live loop, working today: sanitise, triage, synthesise, guardrail, execute, learn.
“We’re cutting the Q3 paid budget by 40% — tell marketing why, get finance to send the final spend number, and draft the replacement plan for review.”
directive · via SlackMessage to #marketing explaining the cut and the reasoning behind it.
Request to finance for the final Q3 spend number, due before the plan review.
Replacement plan drafted and flagged for the requester’s review, with a deadline.
Within seconds of approval the Slack messages exist and the review card has an owner and a deadline. Two days later the Morning Briefing summarises what happened, what’s still open, and why the decision was made — the memory loop, not just the execution loop.
The whole loop end to end, in 70 seconds: watch the demo →
A buyer who feels this weekly, by name.
Operations leads, engineering managers and chiefs of staff at 20–150 person companies running Linear/Jira + Slack, where cross-functional dependencies currently get lost between meetings, threads and tickets.
Focused, not “everyone who manages a team.” This buyer already has budget for coordination tooling — Conductor displaces existing workflow-automation and manual-ops spend rather than creating a new budget line.
Land with one lead. Grow seat by seat.
Conductor lands with one ops or engineering leader routing their own decisions. Teammates start receiving Delegate cards. Then they start submitting their own directives. That’s organic, usage-driven expansion, not a top-down enterprise sale.
Pricing follows the same shape: per-seat for the people who approve, not for everyone the work lands on.
Where it goes once decision volume is real.
Once decision-memory volume is real, the roadmap layers in org-wide rollup briefings, approval policies and cross-team dependency detection — moving upmarket into Ops and PMO.
The long-range destination is full intent infrastructure: the governed plane other agents call through. That’s the five-year story, sold on evidence, not on day one. The sequenced roadmap is on the Architecture page →
The moat isn’t the pipeline. It’s the position.
Being MCP-native is table stakes, and it cuts both ways: the same open protocol that makes Conductor easy to adopt makes it easy to replace. The pipeline is good engineering, not a barrier — structured output is commodity now, and that’s the premise of the whole company.
What isn’t commodity is the position. Conductor is the only layer in the stack with no reason to prefer one vendor’s tool, which is the only place a policy check and an audit trail can sit and be trusted by all four.
Stated as a bet. It becomes a moat the first time a workspace refuses to give it up, and not before.
The compounding asset underneath it is per-workspace decision memory: every approved or rejected card teaches this team’s judgement, and it doesn’t transfer if they switch tools. Whether that produces real retention is what the next eight weeks measure.
What’s de-risked, and what isn’t.
De-risked
- End-to-end execution in four tools today — real tickets and messages in Linear, Slack, Notion and Gmail.
- Zero external sends, payments or irreversible actions without human approval — by construction, not by policy document.
- Per-workspace decision memory live; retrieval under 50ms at current volume.
- The why-now is structural, not narrative.
Not yet de-risked
- Product-market fit. No design partners against the new ICP yet — recruiting five now.
- The moat. Decision-memory compounding is a thesis until retention data exists.
- The value number. We don’t yet know what this costs your team. That’s question one for design partners.
- Team. Solo founder today; one contract engineer is the first use of funds.
The single highest-leverage thing happening at this company is converting the modelled claims into measured ones with real design partners. That is exactly what the raise funds.
What $50,000 buys.
This is a validation bridge, not a seed round. Eight weeks, one question: do ops leads at 20–150 person companies approve what Conductor generates, and does the approval rate climb as decision memory fills?
One contract engineer, eight weeks
Integration reliability and the approval-rate instrumentation.
Design-partner programme
Five workspaces, hands-on onboarding, weekly interviews.
Infrastructure and review
Hosting, model spend, and the security review a 100-person company will ask for on the first call.
What it does not buy: SOC 2, self-serve billing, or paid acquisition. Those are seed-round line items and they are out of scope until the curve exists.
The deliverable at week eight is a one-page proof: approval rate over time across five real workspaces. If the curve is flat, the memory thesis is wrong, and I’ll say so.
Built by one engineer who has been the translation layer.
Conductor is designed and built by Burhan, founder of Min15x. Solo engineer today; the guardrail stage sits in front of execution because a system that acts in your name has to be boring before it can be clever.