Transcription Services

The Vendor is required to provide transcription services to obtain information and recommendations from qualified vendors regarding secure cloud solutions that enable smart storage and retention, audio/video transcription (35+ languages), custom vocabulary, video synopsis, translation (bonus), and training and support—deployed in a dedicated single-tenant AWS gov-cloud environment that meets FBI and CJIS encryption standards and retention controls.
- The solution will primarily be used to transcribe audio files, produce video synopsis, and enforce retention and deletion policies in a manner consistent with law enforcement best practices.
- Requirement:
1. Secure cloud storage
•    Must meet FBI and CJIS security requirements.
•    Ability to set retention policies and delete data when needed.
•    Single-tenant AWS GOVCLOUD account.
•    Smart storage and retention
2. Transcription
•    Convert audio and video to text.
•    Allow custom vocabulary (e.g., narcotics terms).
•    Provide timestamps and speaker identification if possible.
3. Video synopsis
•    Detect objects, places, and actions.
4. Training and support
•    Email support with auto-ticketing and ticket SLAs.
•    Phone support during defined hours; emergency escalation path.
•    Knowledge base (searchable, versioned)
•    Investigator and end-user training (live or on-demand)
5. Security and compliance requirements
•    CJIS alignment: controls and processes consistent with FBI CJIS security policy
•    Identity and access: SSO (e.g., azure ad), MFA, role-based access control (RBAC), least privilege.
•    Logging: comprehensive audit trail (logins, access, changes, exports, deletions), retained per policy.
•    Data residency: all data and processing in gov-cloud (us) regions; no data leaves gov-cloud.
•    Incident response: documented IR process, notification timeframes, and points of contact.
•    Backup and DR: RPO and RTO targets, encrypted backups, restoration testing cadence.
6. Additional and optional
A. Language
•    Support at least 35 languages.
B. Video
•    Ability to train the system with ONIC-specific examples
•    Summarize long videos into searchable clips.