Federal business moved into Teams, Slack, texts, and meeting chats. Under the Federal Records Act those messages are records, but chat platforms are built to forget them. Feith captures every surface with full fidelity and governs it on your schedule.
Agencies adopted Teams, Slack, and texting faster than their records programs could follow. The Federal Records Act didn't change. A message that documents agency business is a record, on whatever app it was sent. The gap between where work happens and where records are kept is the exposure.
The Federal Records Act already covers chat. What fails is the machinery: records tools were built for documents and email, and chat has five structural properties none of them were designed to hold.
Messages edit, unsend, and auto-expire. Native retention often deletes by default, and a user can revise or remove a message after it is read.
A record is fixed the moment it is created. Feith captures each message on a near-real-time schedule, with its thread, participants, and attachments, into a record the source platform can't rewrite.
One decision spans a Teams thread, a follow-up text, a Slack DM, and a line in a meeting chat. No single platform holds the whole exchange.
Records from every surface land in one file plan and correlate into a single case, so the full decision trail survives even when it was scattered across apps.
A flat message export loses what gives it meaning: who was in the thread, what it replied to, reactions, mentions, and the channel it lived in.
Capture preserves the full envelope (participants and roles, thread and reply structure, reactions, mentions, attachments, and channel context) as durable metadata.
Business migrates to Signal, WhatsApp, and personal texting: unmanaged apps where agency records accumulate with no capture at all.
Where the agency authorizes it, encrypted and mobile surfaces are captured through compliant archiving partners and MDM, bringing off-channel business back into the file plan.
Millions of messages, and only a fraction are records. Capture everything and you bury records officers; capture nothing and you lose the ones that matter.
AI triage separates records from transitory chatter. The messages that matter are retained on schedule; the rest expire on theirs.
Each surface is captured through its own native, supported interface: no scraping, no screenshots. Every connector respects the platform's permissions and lands records in RMA iQ with NARA metadata populated on intake.
Capture for encrypted and personal-device surfaces is governed by agency policy and authorization. Feith captures what your authority to operate and acceptable-use policy permit, nothing more.
How the capture layer works ›Feith captures the records an agency must keep, within the boundary the agency sets, and proves it captured nothing more.
Feith captures only the surfaces and accounts your acceptable-use policy and authority to operate permit. Personal devices and encrypted apps are in scope only where the agency explicitly authorizes it.
Capture is aimed at the records of agency business. Transitory chatter expires on its own schedule, and routine personal messages are left alone.
Who captured what, when, and under which policy is logged immutably. The capture program is held to the same accountability as the records it produces.
Capturing the text is the easy part. The work is preserving everything around it: who said it and in what role, what it replied to, the attachments, and the NARA metadata that makes it a record instead of a transcript.
The messages still live in Teams. The record (the full envelope, its NARA metadata, classification and schedule, and a sealed chain of custody) lives in Feith.
The same engine that governs your email, documents, and case files turns captured chat into records you can classify, retain, find, produce, and defend.
Messages are captured in near real time, so the record outlasts the message.
Apply the NARA Capstone approach (comprehensive capture for senior-official accounts, role-based policy for everyone else) or capture in full where the mission requires it. The model is yours to set per source.
CUI categories, sensitivity, and agency markings are applied to messages and threads on intake. AI suggests a classification; a reviewer approves it. Sensitive conversations are confined by policy, not folder convention.
Captured chat inherits the agency file plan. Time- and event-based triggers apply; transitory chatter expires, records persist, and disposition runs through the same multi-stage approval as every other record.
Search every surface at once, by participant, date, thread, or content. Place legal holds that span platforms. Produce conversation records with their full context for litigation and FOIA response.
Every captured message is hashed and sealed. The audit trail is immutable and reproducible: who captured what, when, and the cryptographic proof that the record is exactly as it was when it entered the system.
RMA iQ is DoDM 8180.01–aligned and FedRAMP authorized (JITC-tested under the legacy DoD 5015.02-STD). Access is evaluated per request (ABAC and RBAC). Chat records are governed by the same engine and the same controls as every other record.
The engine behind this: RMA iQ ›Chat records are regulated. Four authorities already define the obligation.
Thanks. We'll reach out shortly.
