Six to twenty-four weeks. Two senior engineers, one-on-one. The thing that's been sitting on your wish list. Your tool, your users, shipped.
Outcomes our archetype students ship
Sally · accountant
“I rebuilt my month-end close as a real tool in week four. My team gets the report Friday afternoon now instead of Monday morning.”
Marcus · ops manager
“Shipped a route-planning tool the dispatchers actually use. Nine ‘wouldn't it be cool ifs’ down to one in production.”
Composite outcomes based on the archetype students described below.
Who this is for
We work with accountants, ops managers, small business owners, founders, students, and side-project builders. Anyone with a friction in their week and the urge to fix it themselves. No coding background required, no minimum ambition. If you have an idea and want to start, you're who we built this for.
Senior Accountant, mid-size firm
“I have an Excel monstrosity I rebuild every month-end. Engineering won't touch it until 2027. I just want it to stop eating my Sundays.”
What we do: Wants to automate a recurring report her team uses. Has tried ChatGPT, hit a wall the moment it asked about a database.
Operations Manager, regional logistics
“I have nine 'wouldn't it be cool if' tools nobody will ever build. I'm done waiting. I just want one of them done so I can show the team it's possible.”
What we do: Has a backlog of small internal tools nobody on the eng team will get to. Wants to ship one and prove the model.
Owner, family-run wholesale brownie company
“I run inventory in a spreadsheet on a laptop I'm afraid to drop. There's gotta be a better way. I'm not going to hire a software firm for this.”
What we do: Runs a small business. Has an idea for an internal tool. Doesn't want to hire an agency, doesn't know where to start.
What you walk away with
We don't hand out completion badges. We help you finish the one thing you've been meaning to build, and leave you with the muscle memory to keep going.
01
Not a demo. Not a prototype that dies on your laptop. Something your team uses, in your real workflow.
02
Your end users touch it from session two, not session six. We design from your grandma's perspective, not the engineer's.
03
You leave knowing which tool to reach for, how to scope the next thing, and how to ship without us in the room.
04
You stop saying "I'm not technical." You start saying "I built that." That part is harder to measure and matters more than the rest combined.
The thing other courses skip
Most vibe-coding training stops at “you built a thing.” That's where the actual work begins. We design from the perspective of the person who has to use what you build: your accounting team, your warehouse staff, someone who's never opened your tool before. If they can't use it, you ship it again.
Step 01
Who is the real user? What is the smallest thing they'd touch this week? Which tool fits you?
Step 02
You build a working version. Could be ugly. By end of session, one real user has it on their screen.
Step 03
Each week opens with "what did your user say?" and ends with "what do they need next?"
Step 04
If someone who's never opened your tool can't use it, we ship it again. We pull on real UX moves, not just the engineer's perspective.
Step 05
You walk away with the tool, the playbook, and the muscle memory to keep going without us.
Three ways to start
Most students begin with the $199 Kickstart, a single session to scope the idea and pick the right tool. If we're a fit, the full $199 credits toward whichever package you upgrade to.
A single session to scope your idea, pick the right tool, and walk away with a plan.
100% credit if you upgrade to a package within 30 days.
Four hours of one-on-one each week, across six weeks. The classic path.
8, 12, and 24-week durations also available. We'll quote yours in the discovery call.
Sixteen hours per week. For the project that needs to ship in weeks, not months.
Best fit for 6, 8, or 12 weeks. Priced for the project, not the hour.
What a week looks like
Four hours of one-on-one with us. Some build time on your own. Real users in the loop the whole way through. Here's a sample week from a Standard student in week three of a process-improvement project.
Always-on between sessions
Day 01
User feedback came in over the weekend. You read it before our first call of the week.
Day 02
We meet for a couple of hours: walk through the feedback together, pick what to change, build the change with your AI tool while we coach.
Day 03
Heads-down day. You keep building or stay on your day job. Stuck moments go to Teams or Slack and we respond same business day.
Day 04
We meet again to ship the change. Your users reload, you watch the moment land, and we zoom out: what did this week teach you?
Day 05
You watch your users actually use it before next week. Notes feed straight into next week's rhythm.
The arc
Whether you pick six weeks or twenty-four, the loop is the same. Longer durations get deeper polish, more user iterations, and more features in your v1.
For students with one specific friction to fix. We pick the tightest possible scope, ship a v1 by week three, polish through week six. Your users have it on their screen for half the program.
Best fit: a single internal tool. Recurring report, one-off automation, internal form replacement.
For students with a bigger project, or for those who finish the first tool and realize the second one matters too. Same loop, just more iterations and deeper polish. Often crosses into vibe engineering territory by the end.
Best fit: a real workflow rebuild, multi-user tools, or building skill across two or three connected tools.
Proof we've lived this
We aren't pretending we invented vibe coding. We learned it alongside the rest of the world. The difference: we did it on top of twenty-five years of real software experience each. Here's what that looks like.
Keith's playbook from this week
A school department needed seventeen multi-page PDF grade-change forms filled out, signed, and submitted as Jira tickets. Took me thirty minutes with Claude Code. Walking through exactly what I told it, what failed, and how to do it yourself.
Read the writeup
How LSA lived this transition
We vibe-coded the first version of our own AI strategy advisor in a weekend. Then we hit the wall every single student will hit. Multi-user state, real auth, embeddings that don't time out. Here is how we crossed it.
Read the journey
Real proof you can keep building
Squawk started as a Friday-night idea. Today it powers our content workflow. Same arc you'll walk: vibe code the thing, hit the wall, learn what to engineer properly. We can show you both sides.
See SquawkMeet your instructors
Senior Advisor · Pacific Time
25+ years building software across enterprise, healthcare, and education. Currently teaches at the university level. The "Sally the accountant" persona on this page is real. That's who Keith trains best, because that's who keeps walking up to him with friction in their week.
This week
Right now: vibe-coding Squawk v2. A content tool that ingests articles, media, and PDFs and builds an automated publishing plan.
Principal · Eastern Time
25+ years in enterprise architecture, government compliance, and AI-product engineering. Has built and shipped production systems with FedRAMP, HIPAA, and FISMA constraints. Mike runs the vibe engineering tier, the bridge from "I shipped a thing" to "my team ships things at scale."
This week
Currently building agent-driven UAT harnesses for our portfolio products.
Things you're probably wondering
Sorted into the two things people actually ask: is this a fit for me, and how does it work with you.
Fifteen minutes. No pitch deck. We'll talk through your idea, the tools you've tried, and whether a Kickstart, a Standard package, or no LSA at all is the right answer.