This tool runs two parallel assessments from your answers: a Robotic Process Automation (RPA) assessment (six dimensions) and an AI agent suitability assessment (four dimensions). Both use the same inputs — no extra questions needed. Each dimension reflects a genuine technical constraint or economic reality, not an arbitrary benchmark. The questions are designed so that the same inputs always produce the same output, giving you a consistent, objective starting point before committing time to a deeper investigation.
Input type & structureThe most fundamental gate. Standard tools cannot reliably read unstructured, non-digital content — this is a technical constraint, not an opinion.
Rule-based natureAutomation executes instructions. If a process requires human judgement that cannot be written as explicit rules, no automation can replicate it faithfully.
StabilityBuilding automation on a process that is about to change significantly is a waste of investment. A standard risk factor in any project scoping.
Volume × handling timeThe standard ROI calculation used in automation business cases. Translates process volume into FTE hours freed — a universally understood metric.
Implementation complexityNumber of applications, virtual machines, OCR, and decision branching are the standard effort factors RPA developers use when estimating projects.
Score weightsBenefit (55%) and ease (45%) are calibrated design choices — reasonable approximations, not a certified standard. The same inputs always produce the same output.
AI Agent LayerEvaluates suitability for AI agents (LLM-based automation) independently of RPA feasibility. Scored on four dimensions:
- Input format — Unstructured or non-digital input scores higher; AI handles what RPA cannot.
- Judgment level — More subjectivity = stronger AI candidate; purely rule-based processes don't need AI.
- Decision complexity — Multi-factor decisions AI can reason through; RPA requires explicit if/then rules.
- Document handling — Scanned documents are an AI strength; RPA OCR is brittle by comparison.