A first engineering role, portfolio-first.

Composite framing. "Devon" is a stand-in for a category of first-role engineer we see often: technically fine, self-taught or bootcamp-trained, running out of savings, and convinced the answer is more applications. The strategist judgment call in every one of these searches is the same — stop applying for two weeks and rebuild the surface. What follows is the shape that intervention takes.

Three hundred rejections and the actual problem.

Devon graduated from a nine-month bootcamp last spring. He has a portfolio site with three projects on it and three hundred applications submitted since April. Two phone screens, zero on-sites. When he lands on Marqee he tells his strategist, Ana, that he needs more application volume.

Ana disagrees before the intake call ends. She's seen this exact shape a hundred times. More applications will not fix a portfolio problem, and Devon has a portfolio problem. Not a technical one — his code is fine. A legibility problem: none of his three projects tell a recruiter what he can actually do in the first fifteen seconds of looking, and none of them look shipped.

The plan Ana proposes is uncomfortable: stop applying for two weeks. Rebuild the portfolio surface. Then re-enter the market with a plan that isn't "spray and pray." Devon takes twenty-four hours to think about it. He agrees. The plan lands the next morning.

Not building new things. Making the existing things legible.

Ana does not ask Devon to build new projects. That would take months. She asks him to make his existing three projects look shipped. This means live demos on a custom domain, three-sentence problem statements above each demo, a screen recording under a minute long, and a README on the top-pinned GitHub repo that a hiring manager could read in ninety seconds and understand what Devon actually did.

Ana writes the first README herself on Monday, in Devon's voice, using a template she's built from reading a hundred senior-engineer README rewrites. Devon uses it as a template and writes the other two by Wednesday. She reviews.

The résumé rewrite happens Thursday. Ana keeps the technical stack and cuts three bullets that read like a bootcamp curriculum ("Learned React and Redux"). Every bullet she keeps starts with a verb, ends with an outcome, and points at a live URL a reviewer can actually click. His résumé is now a single page, and it looks like the résumé of someone who has shipped, because everything on it now points to something you can open.

Building a searchable surface, not a résumé dump.

Week two is LinkedIn week. Ana rewrites Devon's headline from "aspiring full-stack engineer" to something a recruiter would actually search for. She writes an About section that reads like a conversation, not a bio. She uploads a banner. She turns on "open to work" but sets it to recruiters only, not public.

Then she starts outreach. Not at companies — at people. Six engineers at four companies whose stacks match Devon's projects, all reachable at less than second-degree, all with public writing that Devon has actually read. Ana drafts the first three messages in Devon's voice. He edits, approves, sends.

Two of the six reply. One asks for the demo link Devon mentioned. The other, an engineer at a small dev tools company, offers a fifteen-minute call. That call is the first real conversation Devon has had about his work since bootcamp. It goes well. The engineer offers, unprompted, to look at the JD for their open Junior Engineer role and give Devon a warm intro.

The first warm intro, and a hard "no" from Ana.

Jae, the engineer from week two, comes through. She sends Devon a two-line intro to her hiring manager on Friday of week three. Meanwhile Ana has been building a target list of eighteen companies where a Junior Engineer role is either open or was recently filled. Not eighteen open reqs — eighteen companies with signal that a junior seat will open in the next quarter.

Devon wants to blast the whole list. Ana says no, twice. The plan is five outreach threads a week, not fifty applications a week. Anything more than that and Devon can't have a real conversation on the reply side. The math isn't application volume — it's reply-to-conversation conversion. She wants that number above forty percent.

The other landmine of the week is a Discord bot job Devon found through a recruiter friend. The role pays well and would hire him this week. Ana walks him through the trade — taking the bot job now shuts down all eighteen threads in progress, and the bot job doesn't build the kind of résumé he needs for the roles he actually wants in eighteen months. Devon takes the meeting anyway, out of respect for the friend. He passes on the offer.

The strategist wants Devon to send an unsolicited take-home.

By week four, Devon has four active outreach threads and one scheduled phone screen (from Jae's intro). Two of the four threads are cold: an engineer replied but the manager hasn't. Ana wants to break the stall on both.

Her proposed move is unusual. She wants Devon to send an unsolicited take-home — a small demo he'd build in a weekend that solves a problem the target company has publicly said they have. Not a real interview take-home. A gift. Something the manager can look at in fifteen minutes and see him think.

She drops the plan into Devon's approval queue on Wednesday morning with the rationale: for a junior candidate, showing you can scope and ship a small thing without being asked is a stronger signal than a fourth polite follow-up email. The risk is that both managers ignore it. The upside is one of them opens the demo, and that's the interview.

Devon takes forty-eight hours. He builds the demo Saturday. Ana reviews Sunday. It goes out Monday morning. One of the two managers replies within a day. That's the second interview loop.

Two loops, both with technical rounds, both with different flavors.

By week five, Devon has two live interview loops. Both include a technical screen. The dev tools company runs a live coding round with a senior engineer; the second company runs an async take-home followed by a code review conversation.

Ana's prep for the live coding round is not "practice a hundred LeetCode problems." She has him do six problems, spread across three sessions, all in the style the company has publicly said they use. She has him practice talking through a problem out loud with a strategist on Zoom, because half of what junior candidates lose is the vibe of the coding round, not the algorithm.

The behavioral prep is separate. Ana writes out three stories Devon should have at hand: the one where he shipped something small, the one where he got stuck and asked for help, and the one where he changed his mind. That's the trio a junior interviewer is looking for, always in that order.

The take-home prep for the second company is a full weekend. Ana does a code review with Devon on Sunday night, before he submits, and rewrites two paragraphs of the README he wrote for the take-home. Small things. The kind of things a strong senior engineer would notice and quietly downgrade him for if they weren't fixed.

The result isn't an offer yet — it's a real conversation.

Devon's live-coding round at the dev tools company is on Tuesday of week six. Alex, the senior engineer, gives him a small real problem: add a feature to a two-hundred-line Node service. Devon narrates. He gets stuck twice. Both times, he asks a clarifying question rather than guessing. Alex writes in the feedback later that this alone put Devon in the top quartile of juniors he'd screened that quarter.

The take-home for the second company goes in on Wednesday. The code review conversation is Friday. It's not a grilling — it's a pair-programming session with a staff engineer where Devon defends his tradeoffs. Ana had prepped him on the two he'd need to defend the hardest. He defends them cleanly.

By Friday evening Ana writes the debrief. The dev tools company moves Devon to a second round the following week. The second company writes that they'll pass to references. Neither is an offer yet. But Devon has, for the first time in six months, two real conversations in motion — the kind where he's not a résumé in a stack.

What you'd see on your side.

Your tracker looks the way Devon's tracker does. Your first week is a portfolio-and-résumé working session with your strategist. Your outreach threads are small in number and high in effort. You'll see approval prompts where your strategist proposes something unusual — an unsolicited take-home, a direct message where you'd expected a portal — and you'll see the reasoning spelled out before you decide. You're not sending fifty applications a week. You're sending five conversations a week, and each one is one you can actually be present for.