TREAS · TREAS-IRS-56 · record R2842
Accounts Management “Balance Due” Call Transcript Classification
Department of the Treasury · Internal Revenue Service (IRS)
Overview
- Development stage
- PilotSource: b) Pilot – The use case has been deployed in a limited test or pilot capacity.
- High-impact designation
- Not high-impactSource: c) Not high-impact
- Impact justification
- The output is not presumed to be high-impact and is not used as the principal basis for significant decisions/actions
- Start date
- 2024-09-09 (day precision)Source: 9/9/2024
- Withheld from public reporting
- NoSource: a) No
Mission
- Topic area
- Service Delivery
- Operational functions
- Translation and transcription Derived by keyword rules; see methodology
Problem the AI is intended to solve
This project aims to sort customer call transcripts into groups based on the issues customers need help with.
Expected benefits
The AI system will inform research and decision making to enhance the customer experience.
System outputs
The AI system will output categories, classifying call transcripts.
Technology
- AI classification
- Natural language processingSource: Natural Language Processing: AI that processes, interprets, and shares information in human language.
- System name(s)
- Not reported
- Custom-developed code
- Yes
- Public source code
- github.com ↗
Sourcing
- How it was built
- Developed in-houseSource: b) Developed in-house
- Vendor (as reported)
- Not reported
- Vendors (standardized)
- None on the standard list
A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.
Data and privacy
- Involves PII
- Yes
- Privacy Impact Assessment
- Not reported
- Authorization to Operate
- No
- Demographic features
- k) None of the above
- Training and evaluation data
- Two datasets are used to fine-tune the model. The first dataset is the “NXT Switchboard Annotations”. This dataset has 260 hours of speech where researchers manually annotated calls for syntactic structure and disfluencies. Fine tuning is used to clean the call transcriptions. For the second dataset, project members manually labeled a dataset to train a classifier to identify and remove the identity verification portion of the call.
- Federal Data Catalog
- Not reported
Governance
0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.
Pre-deployment testingNo answer reported
AI impact assessmentNo answer reported
Independent reviewNo answer reported
Ongoing monitoringNo answer reported
Operator trainingNo answer reported
Fail-safeNo answer reported
Appeal processNo answer reported
User and public consultationNo answer reported
Potential impacts and how they were identified
Not reported