Munil Shah

We have reimagined the contact center analytics paradigm with the preview launch of the new @Talkdesk 𝗘𝘅𝗽𝗹𝗼𝗿𝗲 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀 interface. Historically, supervisors and operations managers have been forced to navigate complex dashboards, construct manual filters, or wait on dedicated analysts just to extract basic operational signals. The new Explore landing page replaces static reports with "𝗔𝘀𝗸 𝗬𝗼𝘂𝗿 𝗗𝗮𝘁𝗮"—a generative AI natural language interface that dynamically queries end to end contact center metrics via open ended conversational prompts. • 𝗖𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗤𝘂𝗲𝗿𝘆 𝗘𝗻𝗴𝗶𝗻𝗲: Line of business leaders can instantly query metrics across voice and digital channels (such as "What was our service level last month?") and receive real time synthesis without manual report generation. • 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗲𝗱 𝗗𝗮𝗶𝗹𝘆 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀: The platform autonomously calculates daily variance across core metrics, including Total Contacts, Service Level, Abandon Rate, CSAT, Average Handle Time, and Occupancy Rate. It surfaces explicit narrative conclusions rather than raw data points. • 𝗗𝗲𝗲𝗽 𝗗𝗶𝘃𝗲 𝗘𝘅𝘁𝗿𝗮𝗽𝗼𝗹𝗮𝘁𝗶𝗼𝗻𝘀: Users can transition from a high level summary directly into an iterative investigation thread, utilizing the system to instantly isolate anomalous trends or performance changes. By decoupling data synthesis from manual dashboard assembly, we are transitioning contact center data from retrospective reporting to immediate, action oriented operational intelligence.

Patient demand is outgrowing traditional staffing models. In my latest piece for the @Forbes Technology Council ( forbes.com/councils/forbeste… ), I dive into why the shift toward a human and AI hybrid workforce is both a technology and an operational redesign challenge. Traditional IVRs and static bots break the moment an anxious patient deviates from a linear script. Today, agentic AI can handle complex, multi turn conversations. However, most health systems aren't nearly as prepared for this transition as they think because their data remains fragmented across EMRs, CRMs, and contact center tools. @Talkdesk Healthcare Experience Cloud is not just a contact center platform. It is a purpose built vertical solution engineered to securely bridge the gap between clinical data (like Epic and Oracle Cerner) and AI orchestration. By providing deep integration out of the box, it gives AI agents the context they need to be useful while ensuring strict healthcare compliance and governance. Read the full article on Forbes.

🪄 QM Assist + Copilot: Quality Management is getting a nice productivity boost to help QA teams scale their impact across hybrid operations. Admins can now instantly transform manually written evaluation questions into highly optimized, structured prompts tailored for AI scoring consistency. By infusing intelligence into planning and quality assurance, we are making the hybrid contact center more autonomous, and effortless to manage. Get the full breakdown of this week's features here: support.talkdesk.com/hc/en-u…

We continue to expand on our vision for Hybrid Workforce with another important upgrade of our Workforce Management (WFM) and Quality Management (QM) offerings using the power of @Talkdesk Copilot platform. 📝 WFM + Copilot: Resource planners shouldn't spend their mornings buried in granular interval data trying to spot anomalies. Talkdesk Copilot now automatically generates AI-powered text narratives directly on the forecasting page, instantly flagging volume deviations, trends, and anomalies in plain language so planners can optimize hybrid schedules on the fly.

Full release notes here: support.talkdesk.com/hc/en-u…

Get full traceability (workflow, agent, component, channel, version) in one place with the evolution of our AI Conversation Markers report; now unifying Autopilot and AI Agent Interactions. Before this, if something went wrong mid-conversation, pinpointing where and why meant digging across disconnected views. This means faster debugging, cleaner audits, and the confidence to keep expanding your AI coverage without flying blind.

We have rolled out a major upgrade to the @Talkdesk Knowledge Management platform. Our internal benchmark shows significant improvement to the quality and reliability of answers provided by Talkdesk Autopilot and Copilot. Documents up to 50 MB now index end-to-end. Tables (SLA matrices, policy grids, product specs) are fully searchable across keyword, semantic, and RAG. Every connector (internal or external) runs through the same modern pipeline. When Copilot or Autopilot surfaces an answer, you can trust it's drawing from your complete knowledge base, not a partial slice of it.

Full release notes here: support.talkdesk.com/hc/en-u…

WFM Updates: What-If Exports & Queue Management Forecasting hybrid human + AI teams is hard. This release closes the gap with scenario planning, better queues, and advanced AI forecasting.

Studio Switch Contact + E911 Enhancements For the admins keeping it all together: - Real-time CRM identity resolution in flows - New Emergency Calls Report for compliance visibility across regions Built for reliability when it matters most.

Most Frequently Used Email Templates Agents waste seconds hunting for the right reply. Now the most-used templates are ranked and surfaced right in the panel. Small change. Big impact on handle time and consistency.

AI Agent Evaluation & Observability We ship powerful CX AI Agents. Now we’re giving teams enterprise-grade tools to test and monitor them properly. - Realistic eval with regression testing before prod - Full Session History for observability - See exactly how agents behave in the wild Trust at scale starts here.

As CTO, one thing keeps me up at night: AI agents moving fast into production… without enough testing and visibility. Last week’s @Talkdesk releases tackle exactly that. From smarter agent evaluation to everyday agent productivity wins. Quick breakdown 🧵

One of the biggest gaps we see in the market today is that teams are deploying AI agents without a rigorous way to evaluate behavior, reliability, tool usage, and customer outcomes at scale. Nan Zhao from our AI Engineering team wrote a blog explaining how @Talkdesk's latest release for AI Agent Evaluation approaches this problem — from realistic datasets and multi-turn simulations to guardrails, regression testing, and operational metrics that go beyond simple “deflection” rates. We believe it’s among the best AI Agent Eval tools in the market. As AI agents become core to customer experience, evaluation becomes the foundation for trust. Read more: talkdesk.com/blog/ai-agent-e…