← Back to AI Action Plan

How Your AI Action Plan
Gets Built

This isn't a chatbot summarizing your intake form. Your report is built on a structured knowledge base of occupational data, verified tool intelligence, peer-reviewed research, and a multi-step AI analysis pipeline designed to produce specific, actionable recommendations.

Here's exactly what's behind the curtain.

481
Verified research sources
64
Evaluated AI tools
62
Human capabilities scored
17
Industry templates

1Research Foundation

Every assessment is grounded in a curated database of 481 verified sources spanning peer-reviewed research, government statistics, and institutional analysis. This is the same research base that powers jobsdata.ai, a public dashboard tracking AI's impact on jobs, wages, and adoption across 16 prediction categories.

The core finding driving this product comes from a 2026 field experiment by Kim, Kim & Koning (INSEAD & Harvard Business School) involving 515 firms. The study found that companies who received help identifying where AI fit into their operations saw 1.9x more revenue, discovered 44% more AI use cases, and needed roughly 40% less capital to grow — compared to companies with identical tools and training but no task-level mapping.

That task-level mapping is exactly what your AI Action Plan provides.

2O*NET Occupational Task Data

Your task analysis isn't generated from scratch. It starts with structured occupational data from the U.S. Department of Labor's O*NET database — the most comprehensive public taxonomy of work activities in the United States.

We've mapped 30 O*NET work activities across 10 categories to specific AI opportunity profiles. Each mapping includes an AI potential score (high, medium, low), a concrete AI approach, the human judgment that remains essential, which industries it applies to, and estimated time savings.

When you take the assessment, your industry and stated functions are matched against these mappings. The result is an evidence-backed starting point for your task analysis — not a hallucinated list of generic suggestions.

3The 4-Step Analysis Pipeline

Your report isn't generated in a single AI call. It's built through a four-step sequential pipeline where each step builds on the outputs of the previous one. This produces more coherent, internally consistent reports than a single-pass approach.

Step 1: Organization Profile
Analyzes your industry, size, AI maturity, and any uploaded documents to produce an executive summary, AI readiness score with rationale, and quick-win opportunities.
Step 2: Task Analysis
Cross-references your stated functions against O*NET task data and industry templates to identify specific tasks where AI can help most. Each task gets an AI opportunity score, complexity rating, and concrete approach.
Step 3: Tools & Roadmap
Matches your tasks to our verified tool knowledge base of 64 products. Generates a tiered recommendation system (Start Here, Add Next, Consider Later), a 3-phase implementation roadmap, and ROI projections.
Step 4: Risk Assessment
Evaluates displacement risk, skill gaps, change management considerations, and data privacy implications specific to your industry and AI maturity level.

4What Your Report Contains

Your AI Action Plan is a structured document — available as both an interactive web report and a downloadable PDF — containing nine sections:

At a Glance — One-page executive summary with AI readiness score, top 3 actions, projected impact, and the first tool to try.
AI Opportunity Summary — Executive narrative covering your organization's AI readiness, industry context, strengths, and opportunities.
Task-by-Task Analysis — Every key task scored for AI opportunity (high/medium/low), with current process, AI approach, complexity, and expected impact.
Tool Recommendations — Specific products organized in three tiers: Start Here (free/cheap, immediate value), Add Next (compounds gains), Consider Later (higher investment, higher payoff).
Implementation Roadmap — 3-phase plan (0–3 months, 3–6 months, 6–12+ months) with objectives, actions, owners, and estimated investment.
Risk Assessment — Displacement risk, skill gaps, change management notes, and data privacy considerations specific to your sector.
Projected Time & Cost Savings — ROI projections with current cost, projected savings, time-to-value, confidence level, and calculation basis.
Next Steps — Concrete actions to take after reading the report.
Your Inputs — Reference section showing what you told us, so you can verify the report is grounded in your actual situation.

5ROI Projection Methodology

Every ROI projection includes a confidence rating (high, moderate, low) and a plain-language explanation of how the estimate was calculated. We deliberately err on the side of conservative estimates — if the research says 20–40% time savings, we'll cite the lower end.

Projections are built from three inputs: your stated time allocation (from intake), research-backed productivity benchmarks (from our 481-source database), and tool-specific capabilities (from our verified tool knowledge base). For nonprofits and mission-driven organizations, we frame savings as capacity recovered — hours per week freed for mission work — rather than purely dollar-denominated returns.

6Risk Assessment Framework

The risk section evaluates five dimensions, calibrated to your industry, organization size, and AI maturity level:

Displacement Risk
How AI adoption might change roles on your team. Framed as evolution, not elimination — which tasks shift, which grow, which are new.
Skill Gaps
Specific capabilities your team needs to develop to use AI tools effectively. Drawn from our 62-capability framework.
Change Management
Organizational readiness factors: staff anxiety, leadership buy-in, training needs, rollout sequencing.
Data Privacy
Industry-specific considerations for what data can and cannot be shared with AI tools. For nonprofits: beneficiary data, donor obligations, GDPR for international programs.
Common Pitfalls
The mistakes organizations at your maturity level typically make — and how to avoid them.

Risk assessments are industry-aware. A nonprofit gets guidance on board governance, funder perception, and beneficiary trust. A law firm gets guidance on privilege, confidentiality, and regulatory compliance. A restaurant gets guidance on staff communication and customer data handling.

7Verified Tool Knowledge Base

Tool recommendations come from a curated knowledge base of 64 AI and automation products, each independently evaluated and verified. We have no affiliate relationships, sponsorships, or financial incentives tied to any recommendation.

Each tool entry includes verified pricing, specific automation capabilities, integration points, known limitations, and which company sizes it best serves. Tools are filtered to your industry and company size before being considered, and sorted with free/freemium options first for smaller teams.

8Human Capabilities Framework

AI doesn't just automate tasks — it changes which human skills become more valuable. We maintain a framework of 62 professional capabilities across 13 function areas, each scored for how much they appreciate (grow in value) as AI handles more routine work.

These aren't generic “soft skills.” Each capability includes research backing on why it's appreciating, concrete development actions, and specific reasons it resists automation. Your report surfaces the top capabilities most relevant to your functions and industry.

9Privacy & Data Handling

Your uploaded documents and intake data are processed in-memory only. Nothing is written to disk, stored in a database, or retained after your report is generated. Here's the specific data flow:

Uploaded files — Read into memory, PII stripped automatically, analyzed, then discarded. Never touch a filesystem.
Intake form data — Stored in our database to power your report page and dashboard. Contains only what you typed — never file contents.
Report output — Stored so you can return to it. Contains analysis, not your raw inputs.
AI processing — Your data is sent to Anthropic’s Claude API for analysis. Anthropic does not train on API inputs. Your data is not retained by the AI provider.

The research shows that knowing where to use AI matters more than having the tools. This assessment gives you that knowledge.

Ready to find out where AI fits in your work?

5 minutes. Free. Research-backed. Yours to keep.

Get Your AI Plan of Action