Business :: Ideas :: Śavvy AI :: Frameworks :: AI Operational Maturity Model :: Resources :: Zapier's AI Fluency Rubric
⪼ Made with 💜 by Polyglot.
| Level | Description |
|---|---|
| Unacceptable | “This is the extent of how I use AI at work” |
| Capable | “I use AI to operate at a meaningfully higher level.” |
| Adoptive | “I orchestrate AI / build systems that elevate how I work.” |
| Transformative | “I re-engineer how work happens.” |
- Uses AI as a lightweight assist inside a mostly unchanged workflow; helps with snippets, debugging, or summarization but does not materially change how they design, build, test, or ship.
- Cannot clearly explain tool or model choices, limitations, or how their usage has evolved; little evidence of intentional experimentation or repeatable workflows.
- Uses AI regularly across implementation, debugging, testing, and documentation, with concrete examples of better quality, speed, or leverage.
- Shows real tool and model literacy; can explain why they use different tools for different tasks, where those tools break down, and how they have refined workflows over time.
- AI fundamentally changes how they engineer; they default to AI-first approaches where appropriate and have built workflows, tooling, or practices that improve output beyond just themselves.
- Demonstrates strong judgment about tradeoffs and failure modes by choosing tools intentionally, validating outputs, and building review, testing, or mitigation into the workflow.
- Re-engineers how software gets built so AI becomes part of the operating model, not just an individual productivity boost; code production, review, testing, and delivery are meaningfully restructured around it.
- Raises the bar for others by setting standards, building shared systems, and enabling teams to move materially faster without lowering quality, reliability, or safety.
- Uses AI for simple tasks like summarizing, writing, or looking up information, but output reads like obvious AI slop.
- Cannot point to clear evidence that work is faster or higher quality; the way they worked before AI and after AI looks largely the same.
- Has a structured approach for generating specs and prototypes that they reuse and refine across projects.
- Uses AI to tap into user insights (quant and qual) previously inaccessible to the technical limitation, at scale.
- Rapidly generates working solutions users can try and give feedback on; fidelity of what ships to users is both faster and higher quality.
- Can point to entirely new skill sets developed through AI: writing SQL, doing data analysis, building dashboards.
- Building pipelines of agents that take in customer feedback, write specs, prototype, and ship small features.
- Has built always-on AI systems others rely on (e.g., automated monitoring of support tickets, NPS, etc.).
- The PM/Design role on their squad looks materially different than six months ago.
- Can show examples of redesigning how product ships: abandoning previous status quo and reinventing the process.
- Leading new ways of building product, such as unlocking PMs or Designers shipping code in production. Can speak to the tradeoffs.
- Asks AI one-off questions to look something up, then goes back to doing work the same way. AI is a slightly faster Google, nothing more.
- Treats AI as a drafting shortcut for low-stakes written communication (Slack messages, email replies) but hasn’t applied it to substantive work in their role.
- Cannot point to clear evidence that work is faster or higher quality; no examples, no before/after, no signal of impact.
- Has repeatable prompts for core parts of their job. Can describe what they put in, what they get back, and how it’s made that task faster or better.
- Feeds relevant context into AI before complex work rather than asking isolated questions. Understands that output quality reflects input quality.
- Uses AI to do things they wouldn’t have had time to do before: more thorough prep, a second pass on quality, a structured summary for a stakeholder.
- 6 months ago they were doing X manually; now a connected set of tools handles X, and they’ve picked up additional higher leverage work.
- Produces work that others on the team use: a tool, a template, a process. Raising the floor.
- Actively iterates on their AI workflows. They treat their AI setup like a product: it has versions, it gets updated when the work changes, and they can describe what they’ve built next.
- Has changed what their team does, not just how fast they do it. Categories of work either no longer exist or run without human involvement.
- Has influenced how the broader org works: frameworks introduced, processes replaced, ways of working others have adopted.
- Thinks about AI in terms of operating models and role design. Can articulate what roles should look like in 12 months, what will be automated, and what new skills will matter.
- Uses AI for first drafts only. Output reads like unedited AI; hasn’t developed a process for improving quality or adapting tone.
- Uses AI for campaign ideas but has no real workflow. Cannot explain how AI has changed their process, speed, or output quality. Before AI and after AI looks the same.
- Uses AI regularly across content, SEO analysis, and performance review. Output volume and quality are both up, and they can point to specific examples.
- Built a reusable prompt library for top content formats that the team now pulls from. Can explain how they’ve iterated on it over time and why certain approaches work better than others.
- Has run AI-driven experiments with measurable results (e.g., A/B testing copy that increased CTR by 18%) and now defaults to this approach.
- Built a content system that drafts, formats, and schedules posts across channels. The team stopped doing that manually.
- Has built always-on agentic workflows that run without human involvement: content pipelines, monitoring, or campaign ops that operate 24/7.
- Built a personalization engine that serves AI-generated campaign variants at scale, tied directly to pipeline.
- Restructured how the marketing team works: what gets automated, what gets owned, how success gets measured.
- Has automated entire job categories and is driving measurable impact on pipeline and revenue. The team’s output is qualitatively different from what was possible before AI.
- Uses AI summary in Gong after calls.
- Uses AI to help write outreach emails.
- Uses AI to find information for account plans.
- Cannot point to evidence that outreach is more effective, prep is faster, or pipeline has improved as a result.
- Uses AI products built internally daily (e.g., TLU visual generation / Account Prep workflow).
- Built a personal prospecting workflow that saves significant time and generates pipeline.
- Can strongly position AI transformation strategies with customers at a tactical and outcomes level.
- Chains multiple AI tools together into repeatable workflows (e.g., AI research into account plan into personalized outreach sequence).
- Proactively shares what’s working with teammates; has become a go-to resource others learn from.
- Uses AI to surface signals or patterns they wouldn’t have found manually (e.g., expansion triggers from usage data, competitive displacement opportunities from call themes).
- Their AI-driven workflows have produced measurable team-level impact (e.g., significant increase in pipeline coverage, reduction in manual reporting, faster stage progression).
- Actively teaches and enables others; has contributed playbooks, training, or reusable workflows that raised the floor for the whole team.
- Manually does work AI could meaningfully assist with (comp modeling, scenario planning, workforce analysis) and hasn’t tested whether AI would improve it. Skepticism is untested, not informed.
- Actively blocks their team from experimenting with AI. No AI Builder setup, no enablement participation. Becomes a bottleneck for transformation work.
- Uses AI daily across multiple parts of their role with repeatable prompt templates they refine each cycle. Can name the growth arc: started with one-off prompts, now uses AI for every major touchpoint. Quality and speed both improved.
- Has connected AI tools into recurring workflows (e.g., auto-summarizes interview notes, drafts updates for voice). Sets direction for their team’s AI experimentation; carved out time and created safe-to-fail norms.
- Automated a core People process end-to-end with AI-first logic: agent screening, AI-generated deliverables, quality checks that catch issues before they reach decision-makers.
- Replaced recurring manual work (e.g., weekly reports) with live, AI-populated systems. Role has shifted from data compilation to strategic pattern recognition and predictive modeling.
- Stopped running a legacy program entirely and rebuilt the function around AI-first delivery. Agents generate personalized outputs from role data, deliver just-in-time content, and auto-assess completion.
- Redefined team roles around the new operating model; team members now own agentic platforms rather than manual workflows. Successfully upskilled the broader People team and leaders to self-serve within the new paradigm. Produces measurable improvements in performance and effectiveness.