BASE44DEVS

HIRE · VETTING · 6 MIN READ

32-point base44 developer vetting checklist

Vet a base44 developer in 32 verifiable steps across five categories: portfolio, platform-specific knowledge, code samples, references, and contractual readiness. Skipping vetting is the fastest path to a $1,500 sprint that ships nothing. The full checklist takes 60-90 minutes and rules out roughly 70% of marketplace listings.

Last verified
2026-05-01
Category
VETTING
Author
Lead Engineer
Read time
6 min

Why this matters

The base44 hiring market is filled with marketplace listings claiming senior expertise on a platform most engineers learned last quarter. Without a structured vetting process, you will sign a contract with a generalist charging specialist rates and discover the gap at week two. The 32-point checklist below is the verification battery we run on every Base44Devs hire and recommend to teams hiring elsewhere.

Who this is for

  • Hiring managers screening freelance or contract base44 developers
  • Founders who got burned by a previous hire and need a structured next round
  • Procurement teams writing a vendor evaluation rubric
  • Technical leads making the final hire/no-hire call after recruiter screening
  • Operators running a final round on candidates that already cleared a recruiter screen

How to use this checklist

Print it. Run through it in order. A candidate must clear all 32 items at the "must-have" level to be hireable. "Nice-to-have" items are differentiators between two candidates who both clear the must-haves. If a candidate fails three or more must-haves, walk away — they are not the right fit at any price. The red flags page covers what each failure mode tells you.

Section A — portfolio verification (items 1-7)

The candidate must produce evidence of base44 work. "I have worked on base44 projects" without artifacts is not evidence.

  1. Live URL of a base44 app shipped in the last 12 months. Must-have. Must be publicly accessible or accessible via a credential the candidate provides for the call. Screenshots are not URLs.
  2. A second live URL. Must-have. One app is luck; two is a pattern.
  3. GitHub or repository access for one shipped app. Must-have for any engagement over $3,000. Read-only access is acceptable. "I cannot share the code, it is under NDA" is acceptable for one app, not all of them.
  4. A LinkedIn profile with consistent dates. Must-have. Profile dates must match the candidate's claims. Gaps and inconsistencies are red flags.
  5. A portfolio site or case study. Nice-to-have. The case study quality is the diagnostic — vague before/after screenshots without architecture or scope context are a tell.
  6. A non-base44 portfolio piece. Nice-to-have. Demonstrates the candidate has shipped beyond one platform, which matters for migration work.
  7. Public blog posts, talks, or open-source contributions on base44. Nice-to-have, but a strong differentiator. Engineers who teach in public know more than engineers who only do client work.

Section B — platform-specific knowledge interview (items 8-15)

A 30-minute interview covering documented base44 failure modes. The candidate should answer unprompted with specifics, not generalities.

  1. Describe the function-routing 404 failure mode. Must-have. Specialist answer references the routing fix. Generalist answer is "I would check the logs."
  2. Describe credit-burn loops on the platform. Must-have. Specialist references AI-agent re-prompting and loop detection. Generalist has not heard of credit-burn.
  3. Describe the July 2025 SSO bypass. Must-have. Specialist references Wiz's disclosure and the registration-vs-login path divergence. Generalist has no recollection.
  4. Describe AI-agent regression patterns. Must-have. Specialist describes how prompt drift breaks shipped features.
  5. Describe a Stripe-integration failure mode after a platform update. Nice-to-have. Specialist describes the webhook-signature drift pattern.
  6. Describe rate-limiting behavior on base44 functions. Nice-to-have. Specialist knows the per-app and per-user limits.
  7. Describe data-export options off the platform. Nice-to-have, critical if migration is a future possibility.
  8. Describe at least one platform-update breakage they personally fixed. Must-have. Specialist has a story; generalist does not.

Section C — code samples (items 16-22)

A 15-minute review of one code sample the candidate provides.

  1. Code is provided unprompted. Must-have. A senior candidate volunteers a sample without being asked.
  2. Code includes error handling, not just happy-path. Must-have. Look for try/catch, validation, edge-case checks.
  3. Code includes tests or a regression note. Nice-to-have for freelance, must-have for production work.
  4. Naming and structure are consistent. Must-have. Inconsistent naming is the leading indicator of inconsistent thinking.
  5. Comments explain why, not what. Nice-to-have. Diagnostic of seniority.
  6. No obvious anti-patterns. Must-have. Hardcoded credentials, plaintext PII, unsanitized SQL — any of these and you walk.
  7. Candidate can explain the code in their own words. Must-have. Engineers who cannot explain their own code did not write it.

Section D — references (items 23-28)

Two reference checks, 15-20 minutes each.

  1. Reference is reachable. Must-have. If the email bounces or the LinkedIn does not match, the reference is fabricated.
  2. Reference confirms the engagement happened. Must-have. Date range, scope, and deliverables match the candidate's claims.
  3. Reference confirms the candidate shipped on time. Must-have. "Delivered late" is acceptable if the reference explains why and the candidate's account matches.
  4. Reference confirms low rework rate. Must-have. "We had to redo most of the work" disqualifies regardless of the candidate's story.
  5. Reference would hire the candidate again. Must-have. Vague "they were fine" answers are red flags. Specifics or a clean "yes" pass.
  6. Reference offers a third reference unprompted. Nice-to-have. Strong differentiator. References who recommend other references are signaling confidence.

Section E — contractual readiness (items 29-32)

The final five-minute check before the SOW.

  1. Will sign a mutual NDA. Must-have. Refusal is a walk-away.
  2. Will provide a written SOW with milestones. Must-have for engagements over $1,500. See the contract cluster page for SOW structure.
  3. IP assignment on payment, not on delivery. Must-have. Non-negotiable for production work.
  4. 30-day or shorter notice period in any retainer clause. Must-have for agency-shaped work.

Trade-offs and pitfalls

The dominant pitfall is skipping platform-knowledge probes because they feel adversarial. They are. That is the point. A candidate who gets defensive when asked to describe credit-burn is signaling they cannot describe it. The interview is a filter; if it does not filter anyone, it is not working.

The second pitfall is over-weighting portfolio polish. A candidate with a beautiful portfolio site and no platform-knowledge depth is selling design, not engineering. Conversely, a candidate with a sparse portfolio but who can describe four documented platform failure modes unprompted is shipping more than the polished candidate.

The third pitfall is one-reference vetting. One reference is curated. Two is verified. Always insist on two.

How Base44Devs fits in

Base44Devs runs this checklist against every internal hire. We also publish the longer-form vetting guide for teams hiring elsewhere. If you want to skip the hiring loop entirely, order a $497 audit — you get senior platform expertise without the 60-90 day in-house hiring cycle.

QUERIES

Frequently asked questions

Q.01How long does vetting actually take?
A.01

60-90 minutes per candidate for the full 32-point checklist. The portfolio verification (steps 1-7) takes 20-30 minutes. The platform-knowledge interview (steps 8-15) takes 30 minutes. References (steps 16-22) take 20-30 minutes split across two reference calls. Code review (steps 23-28) takes 15 minutes. Contract readiness (steps 29-32) takes 5 minutes.

Q.02Can I shortcut the checklist for small engagements?
A.02

For sub-$1,000 engagements, run only the must-haves: portfolio verification (steps 1-3), one platform-knowledge probe (steps 8-10), and the contract-readiness checks (steps 29-32). That is 25-30 minutes and rules out the worst quarter of candidates.

Q.03What if a candidate cannot pass the platform-knowledge probes?
A.03

Walk away. Platform-specific knowledge is the entire point of hiring a specialist. A candidate who cannot describe credit-burn, function routing 404s, AI-agent regressions, or the SSO bypass under questioning is a generalist who will charge specialist rates and ship generalist work.

Q.04Do I really need two reference checks?
A.04

Yes. One reference is curated by the candidate; you will hear what they want you to hear. Two references — especially if you ask the second to recommend a third — give you cross-verification. The pattern of consistent claims across two independent references is the signal.

Q.05What should I ask a candidate's references?
A.05

Five questions: (1) what was the engagement and outcome, (2) did they ship on time, (3) was rework needed, (4) would you hire them again, (5) is there anyone else you know who has worked with them. Avoid yes/no questions; ask for stories. The narrative quality of the reference's answer is more diagnostic than the score.

Q.06What is the single highest-signal vetting question?
A.06

Show me a base44 app you shipped in the last 12 months that I can actually visit. Live URL, login if needed, and ten minutes to walk through the code. Candidates who cannot produce a live app fail this test. Candidates who can are filtered to the top 25-30% of the market.

NEXT STEP

Order a $497 base44 audit before you hire.

One business day. A senior engineer reviews your workspace, writes a remediation plan, and tells you whether to fix in place or rebuild. Refundable against any fix engagement.