You found out you were bad at the exec update — in the exec update
There’s a moment most engineers don’t see coming.
You spent years getting good at a job with a clean feedback loop. Write code, run tests, ship, watch the graph. When you were wrong, the compiler told you. When you were right, the system worked. The loop was fast, honest, and yours.
Then you got promoted, and the loop disappeared.
Now the work is a budget defense in front of a director who’s already decided. A piece of feedback for a senior engineer who’s brilliant and a little contemptuous. A two-minute update where the VP stops listening at second forty. A reorg you have to announce to people who trusted you. And here’s the part nobody warns you about: you get one take. There’s no staging environment for the conversation that decides whether your project survives. You find out you were bad at the exec update in the exec update.
This is the leap from writing code to running the room. Most engineers make it with no rehearsal at all.
The part of the job that changes everything
Talk to anyone who’s crossed from senior IC into Tech Lead, Staff, or Engineering Manager, and you hear the same things.
The first is identity. You were rewarded, for years, for solving the problem yourself. The whole point of leadership is to stop doing that and make other people successful instead. That’s not a skill upgrade. It’s a different person. Plenty of strong engineers describe the move as a quiet identity crisis, where the thing they were best at is suddenly the thing they’re supposed to let go of.
The second is that the hard part isn’t strategy theory. It’s the human moments. The underperformance 1:1 that isn’t a status meeting. The teammate conflict you can’t refactor your way out of. Feedback that has to be direct without turning into a grenade. People who can design a distributed system in their sleep find these conversations genuinely harder than the system.
The third shows up higher on the ladder, and it’s the most maddening, because it arrives disguised as help. You get told to “be more strategic.” Or to “show more executive presence.” A vague concept gets handed across the desk, dressed up as feedback, and you’re expected to translate it into behavior on your own. Most people can’t. The cost isn’t in that one conversation. It’s in the years that follow, stalling against a standard nobody could quite name. One analysis pegs executive presence at roughly 26% of promotion decisions. More than a quarter of whether you move up, riding on a thing your manager couldn’t define for you.
So you do what engineers do. You read. You buy the books on radical candor and SBI feedback and Staff+ archetypes, you watch the talks, and you still walk into the real conversation cold. Reading about deadlifting is not the same as deadlifting.
That’s the actual problem, and it’s worth saying plainly. Most engineering-leadership failure isn’t a knowledge failure. It’s a rehearsal failure. Engineers don’t lack information about leadership. They lack reps.
Catalyst is a place to take the reps
Catalyst is a private simulator where you rehearse the high-stakes conversation before it happens for real, out loud, against an AI actor who plays the other side.
You pick a moment. The senior engineer who’s curt in code reviews. The PM pushing a date you can’t hit. The headcount you have to defend. Catalyst briefs you on your role, the situation, your objective, and who you’re across the table from. Then you talk. Not type. Talk. The actor stays in character, pushes back, gets defensive, makes you work for it, the way a real person does.
One thing we do on purpose: there’s no score on the screen while you’re talking. Live scoreboards make you perform for the meter instead of staying in the room. You focus on the conversation. The measurement comes after.
When you end the session, Catalyst reads back what you actually did. This is where it stops being a vibe.
Measured, not vibes
Two things get scored today, and they’re the two that quietly decide how you land.
Presence Index asks whether you command the room. It’s built from how you actually spoke: your pace, your pauses, how much you hedge, how steady you stay when the actor pushes.
Decisiveness Score asks whether you make the call. It’s built from commitment language, ownership, and whether you close with a clear next step or trail off into “maybe we could possibly look into it.”
Both numbers come from deterministic rules, not from an AI guessing your leadership potential. That distinction matters. Every point in your report traces back to a specific moment in your transcript: here’s where you softened the ask, here’s where you closed without an owner. The AI writes the explanation. It never decides the number. You’re never asked to trust a black box’s opinion of you.
Then it hands you one drill aimed at your weakest signal, you run it, and you re-measure. That’s the loop: rehearse, measure against the evidence, drill the gap, watch the number move. It’s closer to LeetCode for leadership than to a course you sit through. Over a few weeks the Presence and Decisiveness lines on your Progress page stop being flat. That movement is the whole point.
The honest part: this won’t be comfortable
We’d rather tell you the truth than oversell it.
Your first rep will not feel good. You’ll hear yourself hedge. You’ll see a Decisiveness score you don’t like and recognize, with a wince, that it’s fair. The instinct is to close the tab. Don’t.
That discomfort is the work. You’re not supposed to be good at this yet. If you were, you wouldn’t need the reps. Skill here comes the same boring way it comes everywhere, through patience, repetition, and enough self-belief to sit in the gap between who you are on rep one and who you’re becoming by rep twenty. Nobody gets executive presence from a single brave conversation. They get it from a hundred small ones where the stakes were fake and the learning was real. Catalyst exists so the stakes can stay fake while you build the thing.
Where Catalyst is today, and where it’s going
Honest status: Catalyst is an MVP. Today it does the core loop well. Voice role-play, Presence and Decisiveness telemetry, an evidence-linked report, a targeted drill, progress over time, and a hard privacy line (one human, one AI actor, no employer dashboard, ever). That’s deliberately small, and deliberately real.
Here’s what we’re building toward for V1:
- Tracks with a path, not just one-offs. A 7-day EM interview sprint, a “first 90 days as an EM” track, a Staff+ promotion-narrative track. Structured rehearsal for the moment you actually have coming up.
- A much bigger scenario library, growing from a focused starter set toward hundreds of scenarios, level-aware, each mapped to the signals it trains.
- More signals: deeper voice telemetry like pace and pause analysis, vocal steadiness under pressure, and your question-to-advice ratio, all kept explainable and evidence-linked.
- Personalization, where Catalyst learns your weak signals over time and recommends the next best rep. A custom scenario builder lets you turn a real upcoming conversation into a safe rehearsal without leaking anything sensitive.
- AI-era leadership scenarios, for the genuinely new moments: setting review standards when half the team codes with AI, defending productivity claims you can’t fully measure, navigating uneven adoption.
We’re shipping in the open, and the roadmap will bend toward what early users actually struggle with.
Try it free: 15 days, 2 hours of conversation
You shouldn’t take our word for any of this. Catalyst is in free beta right now: 15 days, up to 2 hours of live conversation. That’s enough to run real reps across a few scenarios and watch your own numbers move. No card, no sales call. Run the underperformance 1:1 you’ve been dreading, read your report, and decide for yourself.
For teams and organizations
If you lead engineering and you’re tired of sending people to a generic leadership course they forget in a week, Catalyst can be tuned to your organization. Scenarios can be shaped to your org’s type, your operating model, and your leadership vision, so engineers practice the conversations they actually have, in the language your company actually uses, against the bar you actually hold people to. This is an organization-level offering: a private practice bench for your engineering leaders, with the same privacy line (no manager-facing transcripts, no surveillance) that makes people willing to be bad at something on the way to being good at it.
Bring your own context
The reps get sharper when they’re yours. Engineers will be able to personalize their training by adding their own materials: your team or org’s vision, your performance-review documents, your yearly professional-development goals. Catalyst can take “improve cross-functional influence,” the line your last review left vague, and turn it into concrete scenarios scored against the behaviors that goal actually requires. The vague feedback finally becomes something you can practice.
The point of all of this
We built Catalyst on a simple conviction. The engineer who can’t get out of their technical lane isn’t lacking talent. They’re lacking a safe place to practice the part of the job that lives outside the code. Give a strong engineer a hundred low-stakes reps at the conversations that scare them, with honest measurement and a clear next step each time, and the ceiling moves. The engineer in training can become the Staff engineer, then the director, then the CTO.
That’s the reason this exists, to make that path rehearsable instead of accidental.
If you try it, we genuinely want to hear what worked and what didn’t. Feedback is the fuel. Tell us where it helped, where it annoyed you, and what conversation you wish it had a scenario for. We’re building this with you, not just for you.
Catalyst — measured leadership rehearsal for engineers. Practice the conversation before it costs you.