# MediaSFU Support: product and documentation summary Canonical website: https://telephony.mediasfu.com/ Provider: MediaSFU Primary purpose: customer-support telephony with human and AI-assisted calling ## What MediaSFU Support does MediaSFU Support gives customer-support teams one workspace for SIP and PSTN calling, operator queues, AI voice workflows, human handoffs, and post-call review. It is designed to keep the customer conversation and its operational context together as work moves between automation and a person. Core capabilities include: - SIP and PSTN inbound and outbound customer calling. - Operator queues, team presence, and call assignment. - AI voice workflows for intake, routing, qualification, and after-hours coverage. - Context-preserving handoffs from AI assistance to a human operator. - Live call visibility and operational supervision. - Call logs, transcripts, notes, and follow-up context. - Workspace, team-access, and calling-configuration controls. MediaSFU Support is a web application. Public product and documentation pages can be viewed without an account. The dashboard and call center require an authenticated MediaSFU account with the appropriate telephony access. ## Recommended evaluation path 1. Review the product overview at https://telephony.mediasfu.com/. 2. Open the documentation hub at https://telephony.mediasfu.com/docs. 3. Follow the user guide at https://telephony.mediasfu.com/user-guide to prepare a workspace and complete a first call. 4. Use the operator manual at https://telephony.mediasfu.com/manual for detailed controls and troubleshooting. 5. Create an account or sign in only when ready to configure a workspace. ## Workspace preparation Before handling a customer call, an administrator should confirm three areas: - Account access: the administrator and operators can sign in with the correct MediaSFU accounts. - Calling readiness: the supported SIP configuration and inbound or outbound routing are available. - Team readiness: operator roles, availability, and escalation responsibilities are understood. Administrators then configure workspace access, connect calling, invite the team, and validate the workflow with a controlled test call. Sensitive provider credentials should remain in the appropriate configuration interface and should never be placed in public documentation or shared troubleshooting messages. ## Operator onboarding Operators should sign in, verify their availability, review the call queue, and confirm that microphone and audio output permissions are working before accepting live work. A short test call should confirm that the operator can answer, speak, listen, use hold or mute when required, and end the call cleanly. When a conversation is handed off from an AI workflow, the operator should review the available intent, account details, notes, and handoff summary before responding. The purpose of the handoff context is to prevent the customer from repeating information already collected. ## First-call workflow A normal customer-call workflow has four stages: 1. Prepare: confirm availability, audio permissions, routing, and the relevant queue. 2. Connect: accept or place the call and verify that both sides can hear each other. 3. Assist: use the available customer context, notes, AI intake, and call controls to resolve the request. 4. Complete: end the call, review the outcome, preserve required notes or transcript context, and identify follow-up work. ## Call audio and media controls Operators can use call controls to manage microphone state, speaker output, hold behavior, and other supported session actions. If audio quality changes during a call, first confirm the selected input and output devices, browser permissions, and network stability. Avoid changing multiple controls at once; make one change, verify the result, and continue. ## Troubleshooting priorities If an operator cannot receive work: - Confirm that the operator is signed in to the intended workspace. - Confirm availability or queue status. - Confirm that the operator has the required role and telephony access. - Refresh live state only after checking whether another active session is using the same account. If a call does not connect: - Confirm the destination or incoming route. - Confirm the supported SIP configuration and provider status. - Check whether the browser, firewall, or network is blocking the connection. - Record the visible call state and time before escalating the issue. If there is no audio: - Confirm browser microphone permission. - Confirm the selected microphone and speaker devices. - Check mute, hold, and system-level audio state. - Test the devices outside the call before changing provider configuration. When escalating an issue, include the time, workspace, direction of the call, visible state, affected operator, steps already attempted, and any safe error message. Do not include passwords, API keys, SIP secrets, access tokens, or private customer information. ## Public documentation map - Product overview: https://telephony.mediasfu.com/ - Documentation hub: https://telephony.mediasfu.com/docs - User guide: https://telephony.mediasfu.com/user-guide - Operator manual: https://telephony.mediasfu.com/manual - Telephony access: https://telephony.mediasfu.com/telephony-upgrade - Main MediaSFU website: https://mediasfu.com/ ## Terminology - Workspace: the MediaSFU Support environment used by a team. - Operator: a person who handles customer conversations and call actions. - Administrator: a person who manages workspace access, configuration, and team readiness. - Queue: the ordered set of customer calls waiting for an available operator. - AI voice workflow: automation used for voice intake, routing, qualification, or coverage. - Human handoff: transfer of an automated conversation to an operator with the available context preserved. - SIP: Session Initiation Protocol, used to establish and manage supported telephony sessions. - PSTN: the public switched telephone network used for conventional telephone calling. This file is a plain-text companion to the public website. The rendered product pages and documentation remain the canonical source.