AI Readiness Scorecard

1. Data Readiness

How would you describe the state of your core data pipelines?

No pipelines — data lives in spreadsheets or siloed databases Basic ETL exists but is fragile or manually triggered Reliable pipelines with some monitoring and error handling Robust, automated pipelines with quality checks and alerting

How clean and consistent is the data you'd use for AI features?

Significant quality issues — duplicates, missing fields, no standards Known issues exist but we haven't prioritized cleanup Mostly clean with documented schemas and some validation High quality with automated validation, governance, and lineage tracking

Do you have enough relevant data to train or ground AI models?

Very little relevant data — we'd need to start collecting Some data exists but coverage is spotty or format is inconsistent Good volume of relevant data, accessible in a queryable format Rich, labeled datasets with clear domain coverage and easy access

2. Infrastructure & Architecture

How would you describe your cloud infrastructure and deployment capabilities?

Mostly on-prem or manual deployments, no cloud-native tooling Some cloud services but deployments are ad hoc Cloud-native with CI/CD, containerization, and basic IaC Mature cloud platform with IaC, auto-scaling, observability, and GitOps

Can your current architecture support async processing and API integrations needed for AI?

Monolithic architecture — adding async or external APIs would be a major effort Some service boundaries exist but integrating new APIs is slow Service-oriented architecture with queues and standard API patterns Microservices or modular monolith with event-driven patterns and API gateway

How mature is your monitoring and observability stack?

Minimal — we rely on user reports to find issues Basic logging and uptime monitoring Centralized logging, metrics dashboards, and alerting Full observability — traces, metrics, logs, SLOs, and anomaly detection

3. Use-Case Clarity

How well-defined is your first (or next) AI use case?

We know we want "AI" but haven't identified a specific use case We have a few ideas but haven't prioritized or scoped them One use case is identified with clear user value and success metrics Prioritized backlog of use cases ranked by impact, feasibility, and data readiness

Have you mapped the end-to-end user workflow for your AI feature?

No — we're thinking about the AI model, not the user experience Rough idea of where AI fits but no detailed workflow mapping Documented user flow with AI touchpoints and fallback behaviors Full UX spec including edge cases, error states, and human-in-the-loop escalation

Do you understand the build vs. buy trade-offs for your use case?

Haven't considered it — assumed we'd build everything custom Aware of some vendors but haven't done a structured comparison Evaluated key vendors and have a preliminary build vs. buy view Component-level build vs. buy decisions made with clear reasoning and cost analysis

4. Cost & ROI Modeling

Have you modeled the costs of running AI in production (tokens, compute, storage)?

No cost modeling — we haven't thought beyond the API pricing page Back-of-napkin estimates but nothing tied to actual usage patterns Cost model exists for POC scale with some assumptions about growth Detailed cost model across POC, 10x, and 100x with per-unit economics

Do you have a clear picture of the ROI or business value AI would deliver?

No — it's more of a "we should be doing AI" feeling General sense of value (e.g., "reduce churn") but no quantification Identified KPIs that AI would move, with rough target ranges Quantified business case with baseline metrics, targets, and payback timeline

Have you planned for cost guardrails and budget controls at scale?

Haven't considered it Aware it's needed but no specific plan Basic budget caps or usage limits planned Token management, rate limiting, caching strategy, and cost alerting designed

5. Team & Process Readiness

Does your team have experience with ML/AI or LLM-based systems?

No AI/ML experience on the team Some engineers have experimented but no production AI experience One or two people with relevant experience; skills gap is identified Dedicated AI/ML capability or strong engineering team with production AI experience

Do you have executive sponsorship and cross-functional alignment for AI initiatives?

AI is an engineering-only initiative with no exec sponsor Executive interest but no dedicated budget or mandate Exec sponsor in place with allocated budget and PM involvement Full C-suite alignment with product, engineering, and go-to-market coordination

How would you describe your team's approach to shipping new capabilities?

Long release cycles, heavy process, change is slow Regular releases but limited experimentation culture Agile with feature flags, A/B testing, and iterative delivery Continuous delivery with rapid iteration, experimentation, and data-driven decisions

AI Readiness Scorecard

1. Data Readiness

2. Infrastructure & Architecture

3. Use-Case Clarity

4. Cost & ROI Modeling

5. Team & Process Readiness

Your AI Readiness Score

Dimension Breakdown

Want a Deeper Assessment?

Contact Info