SmartSimple AI Application Analysis is an intelligent review assistant that uses AI to analyze submitted grant applications and provide reviewers with structured, multi-dimensional insights. By evaluating grant application content across selected fields defined by administrators, the feature generates an overall score, identifies strengths and weaknesses, surfaces flagged items of note, and presents narrative data for reviewers to verify and consider.
AI Application Analysis helps teams focus review efforts where they matter most, accelerating decision-making without sacrificing rigor. The feature is configurable per grant type. Organizations define the criteria that matter most to their priorities and choose which fields to send for AI analysis. Results are retained and accessible over time, allowing reviewers to track how scores change as criteria are refined and grant applications are updated.
Who: Program Officers, Grant Reviewers, and UTA/module Administrators
When to Use AI Application Analysis
Use AI Application Analysis to:
- Reduce the time spent on initial grant application screening by surfacing key insights automatically.
- Sort and prioritize grant applications in the list view by AI-generated score, so reviewers can focus on the most promising candidates first.
- Identify flagged items and weak grant applications earlier in the review process.
- Gain deeper insight into grant application strengths and weaknesses with multi-dimensional scoring and narrative analysis.
- Configure analysis criteria to align with specific program goals and grantmaking priorities.
- Track scoring history over time to understand how grant applications evolve as criteria are refined or as they are updated and reanalyzed.
AI Application Analysis: Privacy, Data, and Design Details
Human in the Loop
AI Application Analysis is designed to supplement human review. All analysis results should be treated as inputs to reviewer judgment and not as final determinations. Reviewer training on interpreting and applying AI-generated insights supports responsible use of this feature.
Large Language Models (LLMs)
AI Application Analysis uses a Large Language Model (LLM). The AI evaluates each grant application fresh against the defined criteria. Each time analysis is run, it reads the selected field content and criteria and generates a response. Criteria can be refined continuously to achieve better results over time.
Data Privacy
AI Application Analysis runs within SmartSimple's own environment and does not send grant application data to external AI services. Grant application data is processed internally.
Language Support
AI Application Analysis currently supports English-language grant applications.
Fields for AI Analysis and PII
SmartSimple allows administrators to choose which fields are sent to the AI for analysis. Administrators can choose to exclude fields containing personally identifiable information (PII).
Data Handling and Location
For the AI Application Analysis feature, relevant data is transmitted from SmartSimple to an internal server via API for processing. The API is deployed regionally, with each endpoint aligned to a specific AWS region. Data for EU clients remains within the EU region throughout the entire process and does not leave that regional boundary.
AI Model
Amazon Nova Pro is the AI model used for analysis.
Software Version Requirement
AI Application Analysis is supported on version 202604.01 or later.
View AI Analysis Results
Once analysis has been configured on a grant application, the analysis runs automatically when the grant reaches the designated status. Click the AI Analysis tab within the UTA/module record to view results. The results include the following components:
Overall Weighted Score
An overall weighted score reflects the combined performance of the grant application across all scoring criteria. The score is calculated out of 10.
Grant Application Summary
The AI generates brief outputs for each grant application: a short narrative summary of the submitted content, and a concise explanation of the score against the defined criteria.
Timestamp
The Processed on timestamp shows when the analysis was last run. Because scores may shift as criteria are refined or applications are re-analyzed, the timestamp helps reviewers know how current the results are.
Scoring Visualization
A scoring visualization displays the grant application's score breakdown compared to the maximum possible score across all criteria. This gives reviewers a visual overview of how well the grant application performed against each criterion and where there may be gaps.
Scoring History
The scoring history view shows how the grant application's score has changed over time. This is useful when the grant application has been updated and re-analyzed, or when the criteria set has been adjusted. Scoring history allows reviewers to compare results across different points in time and understand the impact of criteria refinements.
Flagged Items
Flagged items are issues of note that the AI has identified based on Flag-type criteria. Administrators designate flagged items as things that may deserve a reviewer's attention but are not given a numeric score. Examples of flagged items include geographic location or compliance concerns.
Strengths and Weaknesses
The analysis highlights specific strengths aligned with the criteria defined by administrators. Weaknesses or gaps that reviewers may want to examine more carefully are also included in the AI analysis.
Detailed Criteria Scores
Each criterion in the set displays its results individually, showing:
- The criterion text.
- The score received (for Scoring and Yes/No criteria).
- The weighting applied to this criterion.
- The confidence level for this assessment.
Click View Details to see not only what score was assigned, but why, and how confident the AI was in its assessment.
Resources
For configuration steps, see Configure AI Application Analysis.