The Readiness Gate turns a country's existing forms, guidance, and assessment criteria into a guided digital preparation workspace. Project proponents receive structured, field-level feedback before submission, while the central PIM agency retains control over the standards and final review process.
Project proponents often complete forms without a clear view of the evidence, level of detail, or assessment standard expected. Reviewers then spend scarce expert time identifying avoidable gaps and requesting revisions.
Requirements may be interpreted differently across institutions. Submissions arrive incomplete or weakly evidenced, and formal review begins with repeated clarification cycles.
Expectations are visible while the form is being prepared. Users can identify gaps, improve responses, and reach formal review with a stronger and more consistent submission.
The Readiness Gate sits upstream of formal review. Proponents prepare and improve submissions in the web application, while the central PIM agency defines and calibrates the standards used by the tool.
The readiness result supports preparation. It does not approve the project or replace independent review by the responsible institution.
Line ministries, sub-national governments, state-owned enterprises, and other project sponsors prepare structured submissions with guidance available before formal review.
The responsible agency defines what a strong submission looks like and retains control over the standards, calibration process, and final review decision.
The tool begins with institutional rules rather than a generic AI judgement. These inputs establish the standard the AI is expected to follow.
The structured form proponents complete. A country can use an existing concept note, screening form, feasibility template, or the global template as a starting point.
Explicit standards defining what a sufficient, incomplete, weak, or unclear response looks like for each field.
Practical instructions that explain what information is expected before a response is submitted for evaluation.
The rules and supporting material used to interpret a response in the relevant institutional context.
Previously assessed strong, weak, and unclear responses that allow the agency to test whether the AI applies its standards consistently.
Because the configuration is separate from the application workflow, forms and guidance can be updated as policy changes without redesigning the entire system.
The application gives users a structured path from an initial draft to a stronger submission ready for formal review.
The proposer starts a project using the relevant country form or the global Concept Note template.
The submission is organised into clear sections and fields so the user can work through requirements systematically.
Before answering, the proposer can see what information is expected and how the response will be assessed.
The answer is assessed against the relevant template, guidance, criteria, and supporting policy or methodology context.
The proposer reviews the readiness result, improves the response, and re-runs the evaluation where needed.
The completed workspace can produce a structured report for an existing PIMIS, formal review channel, or other government process.
Evaluation happens at field level so gaps are easy to locate and improve. Every result includes an explanation and practical next steps.
The response contains the information and evidence expected for this field. A Pass is a preparation signal, not project approval.
The response may contain relevant material, but it is not yet clear enough to determine whether all criteria are satisfied.
Required information is missing, insufficiently supported, or inconsistent. The feedback identifies what should be strengthened.
Calibration gives the central PIM agency control over how the Readiness Gate applies its standards. The agency defines the country form, assessment criteria, guidance, supporting context, and reference examples that represent strong, weak, and unclear responses. The AI is then tested against examples that have already been assessed by experienced reviewers.
The tool compares the AI's decisions and explanations with the agency's expected assessments. Where they differ, the agency can review whether the criteria need to be clearer, the guidance or reference examples should be revised, the configuration should be adjusted, or a different AI model should be tested. This cycle is repeated until the feedback is sufficiently aligned with reviewer expectations and approved for use with project proponents.
Calibration does not transfer decision-making authority to the AI. The central PIM agency remains responsible for defining the standards, approving the configuration, and carrying out final project review. The process can also be repeated whenever forms, policies, guidance, or the selected AI model change.
The Readiness Gate supports early preparation. It does not replace formal review, appraisal, or approval by responsible officials.
Feedback is grounded in the country's own forms, criteria, guidance, terminology, and policy context.
Different AI models can be tested and selected according to quality, cost, deployment, data governance, and trust requirements.
Every readiness result includes reasoning and practical suggestions, rather than an unexplained score.
The goal is not to automate project approval. It is to improve the quality, speed, and consistency of what reaches human reviewers.