AI Project Diagnostic
Your AI project is going to fail.
Find out why before it does.
The Premortem is a structured diagnostic that scores your project across six critical pillars and tells you the three things to fix first. No fluff. No frameworks deck. Just a verdict.
Built from patterns across 200+ failed AI implementations. The same six failure modes show up every time.
The Six Pillars
Problem Definition
Is there a real business problem?43% of AI projects start without a measurable pain point. They start with 'we should use AI' instead of 'we need to solve X.'
Data Readiness
Can you actually feed the model?The #1 killer. Teams build architecture before confirming their data is accessible, clean, and sufficient. We score quality, governance, and pipeline maturity.
Workflow Integration
Does it live in a demo or in production?A model that doesn't embed in real daily workflows is a science project. We assess where the output goes and what happens when it's wrong.
Build vs Buy
Are you building what already exists?Teams spend 18 months building what they could buy for $2K/month. We evaluate your approach against commercial alternatives and your team's actual capability.
Success Metrics
How will you know it worked?If ROI targets aren't defined before build, they'll be invented after failure to justify the spend. We check for measurable, pre-committed success criteria.
Adoption & Change
Will anyone actually use it?The model works. Nobody trusts it. No one was trained. The PM left. Adoption failure looks different from technical failure but kills just as many projects.
What You Receive
Readiness Score
Composite score out of 30 across all six pillars
Pillar Ratings
Red, yellow, or green for each pillar with diagnosis
Top 3 Fixes
Highest-leverage changes, sequenced by impact
Cost of Inaction
What continuing on the current path will cost you
Sample Report Excerpt
Meridian Health Systems · AI Triage Routing
Overall AI Readiness: 16 / 30
Verdict: PivotProblem cost identified: $2.1M / year in mis-routes
“Working prototype that nobody trusts enough to use in production. Patient records span three systems (Epic, legacy SQL, scanned PDFs)…”
Common Questions
01How does this differ from hiring a consultant?
+
A typical AI strategy engagement runs $5,000-15,000 over 4-6 weeks. The Premortem is the same six-pillar framework, distilled to the highest-leverage analysis, delivered before your next standup. You get the verdict in 90 seconds for $499. If you need deeper implementation help after, you'll know exactly where to focus.
02What if my project is already in production?
+
Even more useful. Production failures usually happen on adoption, integration, and unmodeled costs — exactly the dimensions this framework checks. Many users come in mid-pilot to find out why traction is stalling. The diagnostic works the same; the fixes are about course-correction, not greenfield planning.
03How is this different from asking ChatGPT?
+
ChatGPT can analyze your project against whatever framework it improvises in the moment. The Premortem applies a specific six-pillar framework derived from 200+ failed AI implementations, with a calibrated scoring rubric and a structured report you can share with stakeholders. The automation is what gets you the verdict in 90 seconds for $499 instead of six weeks for $15,000.
You already know something is wrong.
This gives you the language and the path forward.
Christopher Ferjo
Founder · Previously 8 years at Halliburton · Builds things that work, not things that demo well
15-minute assessment · Report delivered in 90 seconds
vs $50K-500K cost of an AI project failure
7-day money back if no new insights. See a sample report →