Legal, Risk & Compliance · Module
Records Management
Retention schedules and defensible disposition across files, messages and agent memory
Records Management classifies records across One3, applies retention schedules sourced from statutes and policy, and disposes defensibly with approvals and certificates. Nothing under legal hold is ever deleted.
What it does
Records Management is designed to classify records across One3 (files, messages, business objects and agent memory), apply retention schedules sourced from statutes and policy, and dispose of records defensibly with approvals and certificates, while never deleting what a legal hold or other blocker protects.
Classes and schedules
A file plan defines hierarchical record classes, for example Finance › Tax invoices, HR › Payroll or Legal › Executed contracts, each with an owner. Retention rules per class set a trigger (creation, closure, contract end, employee exit), a period, an action (delete, anonymise, archive or review), a statute citation and an entity or jurisdiction scope. A statute citation and source are required. A curated India statute library with sources and verification dates is planned at GA, authored by counsel or a licensed partner.
Classification and computation
Rules on folder, object type and metadata classify records first; AI suggests classes for unmatched files, messages and agent memory items, with low-confidence suggestions queued for confirmation. Retention-until dates are computed from trigger events and recomputed when triggers change; for example, an employee exit event starts retention for linked HR records.
Disposition
Periodic disposition jobs list eligible records and exclude anything under legal hold, open data-principal restrictions, disputes or audits, with reason counts. Class owners review and approve before execution through the owning module's Actions. Each job produces a disposition certificate with class, count, hashed identifiers, rule, approvers and execution time, hash-chained and exportable as PDF.
GA and later
Records declaration locks final records such as executed contracts and board minutes; undeclaring needs approval. An archive tier moves eligible records with checksums preserved. Agent memory and conversation logs follow the same retention and certificate process. Physical records registers are planned for W5.
AI never decides whether a record is eligible for disposition. Planned for W4 (Year 3); retention values must be verified.
Features
What Records Management includes
File plan and record classes
Hierarchical classes with owners and descriptions.
Retention schedules
Trigger, period, action, statute citation and entity or jurisdiction scope per class.
Record classification
Rules first, AI suggestions for unmatched items, human confirmation when confidence is low.
Retention computation
Retention-until dates computed from trigger events and recomputed on change.
Disposition jobs
Eligible records listed, blocked items excluded, owner review and execution via owning modules.
Disposition certificates
Hash-chained certificates with class, count, hashed ids, rule, approvers and time.
Hold and blocker checks
Legal hold, DSR restrictions, disputes and audits checked before any disposition.
Records declaration
Final records locked with versions preserved; undeclare needs approval.
Archive tier
Eligible records moved to archive with checksums and logged retrieval requests.
India statute library
Curated retention rules with sources and verification dates (to verify).
Agent memory retention
Retention classes for agent memory items and conversation logs.
Workflow
Quarterly disposition
- 1
A disposition job lists records past retention
- 2
Records under hold, DSR restriction, dispute or audit are excluded with reason counts
- 3
A summary goes to the class owner
- 4
The owner approves as an R4 Action
- 5
Owning modules execute deletion, anonymisation or archive
- 6
A hash-chained disposition certificate is issued
AI
AI skills and Digital Employees
Record class suggestions
Classes for unmatched files, messages and agent memory within confidence thresholds.
Retention rule proposals
Proposals from a licensed statute library with citations for records manager and counsel approval.
Disposition job summaries
Counts, exclusions and sample rationale for owner review.
Digital Employees that use this module
They act through the same governed Actions as people; risky steps wait for approval.
Frequently asked questions
Can AI delete records?
No. AI suggests classes and summarises jobs. Disposition is approved by people and never runs on held records.
Where do retention periods come from?
From rules with a required statute citation and source, and later a curated India statute library. Values must be verified with counsel.
Does retention cover AI agent memory?
Yes. Agent memory and conversation logs carry record classes and follow the same disposition process.
See Records Management in your workflow
Tell us about your team and the outcomes you want. We will walk you through the apps, the Digital Employees and the plan for your industry.