The Form Fires Four Things at Once. Only One of Them Can Fail.

We launched the AI Operations Academy this week. It's our structured training arm: half-day workshops, executive strategy sessions, and an ongoing enablement partnership for service companies that want to use AI without hiring a data science team.

The front door is a $750 AI Readiness Assessment. Not a free download. Not a lead magnet. A paid, scoped discovery session that anchors the conversation before anyone commits to a $3,500 bootcamp.

That choice is deliberate. Free lead magnets attract browsers. A paid assessment attracts people who've already decided to move. It filters the pipeline at entry rather than at close.

What Happens When the Form Submits

The assessment form collects name, email, company, and an optional team context field. When it posts to /api/academy-assessment, four things happen at the same time:

1. A confirmation email goes to the person who submitted, telling them Isaac will reach out within one business day.
2. An internal notification goes to our inbox with the full submission.
3. A lead record lands in Airtable via our shared captureLead utility.
4. A PostHog event fires server-side to track the conversion.

We use Promise.all to run 1, 2, and 3 simultaneously rather than chaining them sequentially. On a cold Vercel function, each of those calls can take 300ms to 500ms. Sequential would mean 1.5 seconds of wall-clock time waiting on three independent network calls. Parallel brings it closer to the slowest single call.

Error Priority

The four operations are not equally important.

If the confirmation email to the person fails, we return a 502 to the client. The form shows an error. They can try again. We'd rather surface that failure than silently succeed while the person waits for a confirmation that never comes. A lead who doesn't receive a confirmation email assumes it didn't go through and may not trust the follow-up when it arrives.

If the internal notification fails, we log it and continue. Isaac can find the Airtable record. The lead isn't lost.

If Airtable fails, we log it and continue. The email is already sent. We have the submission. We can reconstruct the record.

PostHog fires after the response is returned. If it fails, it fails silently. Analytics are not a dependency of the user experience.

The code makes this explicit:

const [userResult, notifyResult, leadResult] = await Promise.all([
  sendUser,
  sendNotify,
  recordLead,
]);

if (userResult.error) {
  return NextResponse.json({ error: "Failed to send confirmation" }, { status: 502 });
}

if (notifyResult.error) {
  console.error("[academy-assessment] Notification email failed (non-fatal):", notifyResult.error);
}

The user email is the critical path. Everything else degrades gracefully.

The Three Programs

Once someone completes the assessment, the conversation branches into three tracks:

The AI Operations Bootcamp ($3,500 for up to 15 people) is a half-day on-site or remote workshop. The guarantee is simple: three deployable workflows leave the room with your team, or you get a full refund. We capped it at a half day on purpose. Full-day trainings run out of gas by hour five. A half day keeps people sharp and ends before the energy drops.

Executive AI Strategy Sessions ($2,500 per session, or $7,500 for three) are 90-minute deep dives for owners and leadership. Tool selection, vendor evaluation, a 90-day integration roadmap. Built for the person who needs to make decisions, not the team that executes them.

The AI Enablement Partnership ($4,000/month) is ongoing. One workshop a month, async support for every question the team hits in between, and quarterly audits of what's actually been adopted versus what sounded good in the room.

Why the Assessment Comes First

We tried leading with the bootcamp offer directly. The conversations were longer and harder to close. Someone who inquires cold doesn't know what they need. The assessment creates a shared foundation: we understand their team, their tools, their actual bottlenecks. They understand what the training is designed to do.

The $750 also frames the relationship. This isn't a free audit we're giving away to earn a sale. It's a deliverable. When it's done, they have a report, a prioritized recommendation, and a clear picture of which track fits them. Whether or not they book a program, they've gotten something useful.

That framing closes faster than "let us tell you about our training." It also surfaces the people who are actually ready versus the people who are still deciding whether AI is relevant to them.

The pipeline is live. You can book an assessment at twenty1-media.com/academy.