AI Agent Portal, AI Course Generation, and Real Call Data in Training Scenarios
Ask growth questions in a chat interface, get follow-up emails drafted for your approval, and build a full course from one topic. Here’s everything new.
Coaching Programs, Mobile Practice, and Sharper Weekly Digests
Coaching programs, practice from a phone browser, and a weekly digest that names the two criteria each agent needs work on rather than offering a generic tip.
Behavior Evidence, Reliable AI Coaching, and Mobile Scorecards
This Week’s Highlight You can now click any detected behavior from the dashboard or a rep’s profile to see the exact transcript lines that support it. Behavior detection is also more precise, with stronger alignment between behaviors and the relevant skill and less noise from routine call greetings. Also This Week AI Coach sessions are more reliable The coach no longer repeats the same question in a loop, and skipping a question now moves the session forward without preventing the training from being created. Sessions open reliably, conversation context is retained throughout, and persona portraits now appear instantly instead of occasionally loading blank. Mobile scorecards now score the right person Post-roleplay evaluations on mobile now score the trainee’s responses rather than the AI persona’s lines, so scorecards accurately reflect your practice performance. Scorecards also now render correctly the first time instead of occasionally appearing blank until the page is refreshed. Gaming alerts are firing again and are now available via API We fixed an issue that prevented gaming alerts from triggering consistently. Gaming alert data is also now available through the developer API, making it easier to surface alerts in your own systems without logging in to Insight7. Trainings navigation is now consistent and predictable All account types now land in the same Trainings destination, and roleplays are listed in priority order. We also fixed the back-button loop that could bounce users between a scorecard and the assignment page. Fixed & Improved Ready to put the coaching improvements into practice? Open Insight7 Skills Practice & Roleplay and start a training session.
Behavior & Skill Intelligence, Training Deadlines, and Mobile Roleplay
This Week’s Highlight Behavior and Skill Intelligence is now available for organizations with access. It detects observable behaviors across every call, scores them independently of your existing criteria, and rolls them up into skills at the agent and team level. Results now appear in a dedicated scorecard tab, on agent profiles, and on the team dashboard. The What’s driving your outcomes section also surfaces specific, data-backed insights into the behaviors and skills influencing performance. Also This Week Set fairer deadlines for training assignments When assigning a training, you can now choose a completion window in days. Each learner’s due date is calculated from their individual assignment date, so staggered cohorts get the same amount of time to complete their training instead of working toward one shared calendar deadline. Due dates appear on the learner’s training list, the pre-session screen, and the assignees view. See who started a training but hasn’t finished The training dashboard now gives you a clearer view of learner engagement. A new Created column shows when each training was created, and the table now sorts by most recent by default. There’s also a new Attempted column that distinguishes learners who opened a training but didn’t finish from those who haven’t opened it at all. The Attempted column replaces the previous Improvement column, making it easier to identify who may need a follow-up. Key pages now load much faster We’ve improved performance across several frequently used areas of Insight7. People search, the evaluations list, the training dashboard, and the transcripts view now load significantly faster, helping teams get to the information they need with less waiting. A redesigned mobile roleplay experience The mobile roleplay experience has been redesigned to give learners more flexibility and context before and during practice. Learners can now choose from six built-in scenarios or create their own, move through a scene-setting step before each session, and receive richer feedback during the experience. Authored lessons from learning plans also now support the full block-based content experience on mobile, including progress tracking. Fixed & Improved If training assignments or coaching tips are part of your team’s regular workflow, it’s worth a quick check this week to confirm your settings reflect how you want them to run.
Weekly Lowest-Score Digests, Daily Mobile Practice, and Smarter Gaming Alerts
This Week’s Highlight Area managers can now receive a weekly email listing their reps’ lowest-scoring calls, complete with links to scorecards and coaching tips for each one. The digest is fully configurable: set your score threshold, minimum call length, send day, send time, and timezone from Settings under Evaluation Alerts. Changes take effect on the next scheduled send with no extra steps needed. Also This Week Practice speaking skills every day, right from the mobile app The home screen now shows a Daily Activities section with two scored speaking drills: Hot Mic for open-ended takes and Word of the Day for prompt-driven responses. Both are recorded, scored, and saved so you can revisit them. Five Speech Games (Word Chain, Say It Shorter, Change Direction, A Strange Question, and Pick a Side) are also now fully scored with session history. The gaming alerts dashboard now shows accurate results and lets managers take action Managers can also now mark any alert as resolved directly from the dashboard. Your organization’s timezone is now one consistent setting Timezone used to be stored in two separate places, which could fall out of sync and silently default to the browser’s local timezone. There is now a single timezone per organization, set intentionally from Settings. The field starts empty until an admin confirms it, so if your organization had not explicitly set a timezone before, check Settings to confirm the correct one is saved. This affects training reminder schedules and the weekly digest send time. Fixed & Improved If your team is ready to put the new weekly digest to work, ask your account contact to enable it and then head to Settings under Evaluation Alerts to configure it for your organization
Live Chat Training, Faster Background Evaluations, and Pidgin English Voice Personas
Chat training now streams and branches in real time on the same engine as voice, evaluations process in the background for faster results, and voice personas can now speak Pidgin English.
Smarter Team Invites, Custom Training Reminders, and Flexible Speaker Matching
This Week’s Highlight The team invitation experience has been completely rebuilt. Clicking an invite link now opens a dedicated acceptance page where new teammates can set up their account before entering Insight7. The new flow also prevents automated email security scanners from invalidating an invite before the intended recipient opens it, making team invitations smoother and more reliable from start to finish. Also This Week Agent profiles are now created automatically When you invite someone with the agent role, their evaluation profile is automatically created as soon as they accept the invitation. Admins no longer need to complete a separate setup step before a new agent can access their calls and evaluations. Google Drive imports are more reliable We’ve improved the Google Drive experience across the full connection and import flow. The file picker now closes properly after a file is selected, authorization returns you to the page you started from instead of redirecting you to the dashboard, and temporary network issues during authorization are handled more gracefully. Calls imported automatically from Google Drive now also trigger Slack signal notifications, matching the behavior of manually uploaded calls. Set your own training reminder schedule Training reminders are now fully customizable from Notifications settings. When reminders are enabled, you can choose how many nudges to send, the number of business days between them, the send time, and the timezone. A live preview shows exactly how your reminder schedule will work before you save it. Choose flexible or strict speaker matching for evaluations Teams can now control how agent names in transcripts are matched against their roster. Flexible matching, which remains the default, uses approximate name matching as before. Strict matching only accepts exact name matches, giving teams more control when precision matters more than coverage. Existing evaluation results won’t change unless you update the setting. Fixed & Improved
Evidence-Backed Agent Scorecards, Personalized Roleplay Training, and Workflow Signals
This Week’s Highlight Agent scorecards are now powered by real evaluation data, giving supervisors and agents a clearer picture of performance in one place. Every assessment includes an overall score, a breakdown by criterion and sub-criterion, evidence quotes pulled directly from the conversation, a narrative summary, and coaching tips for each area. This makes it easier to understand not just how an agent scored, but why—and what they can work on next. Also This Week Low satisfaction scores now automatically route cases to a recovery team After a case is resolved, the customer’s satisfaction response is captured and the case is routed based on the result. A positive score closes the case as usual. A poor score moves it to a dedicated Recovery tab, where the team can review the case and close it out. The satisfaction rating is also visible directly on the case. Agents get personalized roleplay training generated from their QA results When an evaluation identifies areas where an agent scored below expectations, Insight7 can automatically create and assign a targeted roleplay training session based on those results. The agent completes the training as a live voice session and receives a scored results screen afterward, connecting QA findings directly to practical coaching. A new Customer Experience dashboard with Workflow signals is now available The updated Customer Experience dashboard includes a new Workflow signals tab with busyness and trajectory breakdowns, a ranked pipeline signals list, and a per-call workflow table. Evidence quotes link directly to the source transcript, while each category includes a concise summary alongside the supporting evidence. Evaluation filters now carry over to the call evaluations list Filters applied on the main dashboard now remain active when you navigate to the call evaluations page, so you no longer need to recreate the same filter setup. Fixed & Improved
Voicemail Filtering, Smarter Search, and More Realistic Roleplay
This Week’s Highlight Agent scores are now cleaner and more trustworthy with the new voicemail filter. Managers can view all calls, show only voicemails, or exclude voicemails entirely from the evaluation dashboard and agent profiles. Voicemail detection happens automatically, and the filter works with calls that have already been scored—not just new ones. That makes it easier to evaluate agent performance based on the conversations that matter most. Also This Week Find the right recording faster with redesigned search Title and keyword search are now combined in one place with a simple toggle, making it easier to find the conversation you’re looking for. Search results now show how many times your phrase appears along with a short excerpt around the first match, so you can judge relevance before opening a recording. Recent searches are also saved automatically in your browser. Practice with more realistic roleplay personas Roleplay coaching is now more challenging and more natural. In guided discovery trainings, personas no longer volunteer information the trainee is expected to uncover. Instead, they reveal relevant details as the conversation calls for them, respond according to what they know, and adapt their reactions as the session unfolds. The result is a more realistic practice experience that rewards strong questioning and discovery skills. Practice everyday conversations with personal training scenarios Training is no longer limited to workplace conversations. A new Personal category supports practice for small talk, social situations, family conversations, and personal negotiations. These scenarios are designed around peer-to-peer conversations rather than traditional buyer, customer, or workplace dynamics, with feedback tailored to the situation. See consistent all-time scores on agent profiles Agent profile scores are now clearly labeled and consistent. The score displayed on an agent’s profile is pinned to their all-time overall score, so it no longer changes silently when you adjust the date filter. Fixed & Improved
