Intern conversion as the dominant full-time hiring channel

Outcome: Run a real internship program as the primary full-time hiring channel — three months of real work is the highest-signal trial available.

Context: Galane "just launched internship right now" for exactly this reason — Chandler himself interned four times at SpaceX before going full-time. Critical work on Starship/Dragon/Falcon is largely done by intern conversions.

“the, the, the value that I can add here is particularly related to the internship funnel. I think given the brand with Tesla and SpaceX, like literally it's been quantified. These are like the most applied to programs ever... it gives, it gives a three month trial period to these people Yeah. And people that crush stay all the time... the overwhelming population of folks that are doing so much of the critical work on Starship Dragon, Falcon even are intern conversions.”

Democratize internal data — kill access controls and put every decision history in one searchable backbone

Outcome: Build a single internal data backbone with effectively no internal access controls so every engineer can see the full context and history of every decision.

Context: Implementation: web-app-hosted core engineering information, no hard-drive silos, no email-thread context, integrated data frame storing decision history, LLM layered over the repo so folder-structure ignorance doesn't block navigation.

“everything is like web app posted. Their access controls are basically gone, at least internal to the company when it comes to being able to see the core engineering information. It does not live on a hard drive somewhere. It does not live in an email that gets sent to a specific group of people... We've really focused on like building that integrated data frame that enables anyone to access and see the context of like why decision was made or what decision was made... now with LLMs you can kind of like let them run wild on top of your data repository and ensure that if someone doesn't understand the folder structure, they can just query the LLM”

Pull in subcomponents to in-house manufacture only where they bottleneck the production line

Outcome: Identify the 3–5 assemblies that bottleneck your end-state production rate and pull those in-house first — leave the rest with suppliers.

Context: Galane's target: 10,000 missiles per year. The five "screaming" subassemblies are the active candidates for in-house manufacture, with weldments as the leading example. Capital and time analysis comes second to the bottleneck-identification step.

“we have to bring in the assemblies in-house that, that are going to bottleneck our supply chain as fast or bottleneck our production line as fast as possible. So I think what that looks like for us is, is some of the bigger weldments, like starting, starting with that sort of thing, bringing in-house sooner rather than later is something that's relatively easy to do but is a very complex thing that has multiple steps... bring that in-house, we, we, we whack one mole on the table of of, of getting to this, you know, 10,000 missiles per year number. I think there's of course many more challenges or you, you kind of just, whenever you're looking at your in-state product, you need to go and see what are the things that are, are really hitting me on schedule. And for us, there's probably five of those right now that are, are like screaming at us. Like, please help, please help.”

Multi-screen technical interview + behavioral panel for full-time hires

Outcome: Run 6+ technical conversations + a tech test + behavioral panel before any offer, and treat the slowness as a feature.

Context: Mariana mirrored SpaceX/Tesla's depth — 8–10 conversations before offer — even though the brand draw is smaller. Pitch the rigor in both directions: candidate also gets to assess the team they'd be embedded with.

“the like technical evaluation. It goes quite deep. And so they're, you know, you're going to talk to six engineer if you're applying for an engineering role, you're probably talking to six engineers before you're getting an offer. You're almost certainly doing a technical test that kind of shows how you think through problems and are you able to solve the problems that your resume says you're able to solve. And so that level of like technical rigor that goes into assessing folks that are coming into the company is pretty extensive. You, you're gonna have eight to 10 conversations before you get an offer.”

Set the schedule by sitting down with executors and battling it early

Outcome: Schedule-setting is a sit-down with executors where the CEO defends the timeline against the person who has to ship it — then lock it in early.

Context: Galane sat down, set the ambitious "rocket in the air by June" schedule, broke down the tasks, and any subsequent drift triggers a "what's wrong" investigation.

“doing something sort of upfront to align the team to like, Hey, this is what I, this is how long I think these things are gonna take you as the person who's gonna go execute on it directly. Like, does this sound reasonable? And then battle it there early and then set that schedule from the get go, which is what we did. We all sat down, we set this very ambitious schedule to go get a rock in the air by June. And like we, we broke down the things we need to do to get there”

Founder personally interviews every hire, often multiple times

Outcome: In sub-50 stage, founders personally interview every hire and re-engage them multiple times to close on mission.

Context: Both founders confirm this practice as still active. The repeat calls are partly "coercion" — selling the mission while assessing fit.

“we're still very small, so it, it's, it's not really saying much, but I'm of course talking to every person. I, I don't know if you still are. I'm I'm sure you are. Multiple times. Exactly. Yeah. Multiple times. Coercing them over the phone, convincing them that”

Recruit university project teams as units, fly the whole team to your city

Outcome: Identify university project teams that use your exact technical stack and recruit the whole team to a single physical location.

Context: Chandler identified a specific (non-US) university rocket team using Galane's propellants; the plan is to fly the team to Austin together. Team chemistry transfers; passion is pre-filtered.

“I actually found specific rockets, even the country that that works on ro rockets with the same propellants that we're easing, which I did not know existed. cause it, it's a little bit more rare, but I'm targeting that school. Get them all out, get them all out to Austin, get the whole teams, get em on a bus and bring em over here.”

Structure initial team by discipline so critical-path swarming doesn't pull avionics off engine problems

Outcome: Structure the initial 6–20 person team strictly by domain/discipline so critical-path swarms stay on-discipline.

Context: At six people, this is easy to control; the founder is calling it out as top-of-mind for the growth-team phase because losing it wastes resources fast.

“we have right now our team is six people, so it's, it's relatively easy to to control that game. But, but it's still very important. Like, you know, how how we've structured all of the initial team is, is disciplinary based or, or sort of domain based. So it may not make the most sense for a avionics engineer to go troubleshoot a engine design problem that's literally blocking the position.”

Auto-generate shift pass-downs from the data backbone — humans review, edit, click send

Outcome: Auto-populate daily shift pass-downs from the shared data backbone so humans only review, edit, comment, and click send — keeping accountability without the writing tax.

Context: Tesla observed even large operations write pass-downs manually. Mariana built data backbone instrumentation so end-of-day reports are generated, then human-owned by review-and-send.

“because that does become burdensome from like a, as like the team starts to grow and as there's a lot of stuff happening, we've done our best to try to, like, if everything is going into the same, like aggregated kind of like data backbone, you can actually auto-populate the bulk of those pass downs and it puts humans in the position where they're really like reviewing a pass down, editing, Maybe adding, adding commentary versus having to write something from Scratch... you still want people to look at it. You still want them to click send because like, they should have ownership and accountability of like what is in it.”

High-cadence email updates from the SWAT-team owner during critical-path execution

Outcome: The critical-path SWAT owner sends high-cadence, high-signal email updates to the full company throughout the push.

Context: Chandler keeps a notebook in his pocket at all times. The owner writing it down forces them to confront "I don't think I made the most direct progress today, let's fix it tomorrow."

“high signal, low noise email updates on, particularly with things related to critical path are, are extremely important... for the, the, the fire teams, there's usually gonna be one extreme owner who is yeah. Driving that problem or driving that, you know, specific task to get it completed. And for that person to take ownership and send, you know, high, high cadence email updates to get through the not so good parts of the said problem is, is extremely important not only for the team to see and in hear, but I think it's actually wildly important for the individual who's working that project to recall what they, what happened that day. Write it down”

Set company-wide aggressive milestone, decompose into tasks, then sort what doesn't fit

Outcome: Set the aggressive milestone, decompose into all tasks, then attack-or-delete the ones that don't fit the timeline.

Context: Attack means focused SWAT; delete is ideal. "Ideally float those things to the top, ideally delete." This is how aggressive timelines translate into focused execution rather than panic.

“It's like, okay, there's a thousand things that have to happen if we wanna do it in six months. 900 of those things can be done in six months, but a hundred of them can not be done in six months. And so we have to go attack those hundred things. It's attacking Function attack can mean delete them too ideally. Exactly. Ideally float those things to the top, ideally delete and they cut Did. That's right.”

"Walk me through a problem you solved" — single broad question + 15–20 minutes of jab-and-punch

Outcome: Open every technical screen with one broad "walk me through a problem you solved" and probe depth for 15–20 minutes — signal arrives quickly.

Context: Chandler reports the answer is clear inside 15–20 minutes regardless of the candidate's discipline; cross-discipline probing reveals reasoning structure.

“I always start with one sort of broad question and dig into a lot of things, but I I, I really open it up to, to allow myself a playground to like punch and jab and, and go around their problem set. But I, I effectively ask, walk me through a problem that you solved. And whenever they walk through that problem over 20 minutes or 15 minutes, I, I know very quickly whether or not they're good or not good. And usually I'm, I'm okay enough no matter what the discipline is to kind of jab and punch around that problem set a little bit to, to feel it out.”