Once your transcript files have been successfully uploaded and processed through the Transcript Reader, you will receive output files via SFTP. These files contain critical results, diagnostic information, and additional metadata about the transcripts.
Output files overview
After processing, you will receive the following files - please note, each output file will contain reference to a version id. As the Transcript Reader continues to be enhanced, the version id may be updated:
1. GPA results file ([school name][date][batch ID].transcripts _results_.csv)
This file provides GPA information for each transcript. It includes multiple types of GPA calculations, allowing for different interpretations based on school or program-specific methodologies.
What’s inside:
Extracted student name, high school
Extracted GPA Values (4pt and 100pt)
Multiple GPA calculations (e.g., weighted, unweighted, core GPA)
Course Counts (e.g AP / honors)
For more details:
Refer to the article on GPA Calculation Methodologies to understand how each GPA value is derived.
2. Application results file ([school name][date][batch ID]. applications_results_.csv)
This file provides GPA information for each application. If a student has multiple transcripts, these have been aggregated to provide an overall GPA calculation for the application.
You may notice that several columns may be blank - this file does not include any extracted values from the transcripts (i.e weighted and unweighted GPA). This is because this file represents results at an application level, and if an application has multiple transcripts, this would result in several extracted GPAs on the transcript.
For more details:
Refer to the article on GPA Calculation Methodologies to understand how each GPA value is derived.
3. Error File ([school name][date][batch ID].row_errors_.csv)
This file includes transcripts that could not be fully processed and require human review.
Reasons a transcript might appear here:
Extraction Errors: Indicates the reader was unable to extract necessary data from the transcript. This may be due to transcript format / quality of the pdf
Calculation Errors: Data was extracted but failed our accuracy benchmarks or produced an error during GPA calculation (e.g., missing credits or invalid grade formats).
Low Confidence: Indicates the reader has low confidence in its ability to compute an accurate GPA, due to various risk factors (e.g. abnormally low GPA, potentially incomplete extractions, significant misalignment between the raw GPA printed on the transcript, and the computed GPA, and/or high credit variance between school years).
Action Required: Review the flagged transcripts, and calculate the GPAs manually. These transcripts should not be re-submitted for processing.
4. Additional Transcript Information File ([school name][date][batch ID].full_results.json)
This file provides enriched data extracted from the transcripts for deeper analysis or integration into downstream systems.
Current contents include:
Course level metadata:
Difficulty (e.g., Honors, AP, IB)
Credits per course
Grade earned
Academic year (e.g., 10th grade, 11th grade)
Term details (e.g., Semester 1, Fall 2023)
High school name
This structured data is useful for custom analytics or to power advanced AI processing pipelines.
5. File Errors ([school name][date][batch ID].file_errors.csv)
This file indicates any transcripts which the AI Transcript Reader is unable to process due to the file format. Common reasons for this are as follows:
The directory has been zipped (instead of just the folder)
There is misalignment between the pdf file name listed in the Index.csv and the pdf's themselves
The Index file is a .txt or .tsv format, instead of civ
The Transcripts are not in pdf format
Action Required: Review the transcripts in this file, make the required adjustments and re-upload to the transcript reader for processing. This file will appear in the SFTP shortly after the transcripts have been uploaded.
For any questions and concerns, please reach out to the CollegeVine partnerships team!