<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Agentic Index: voice agents change log</title>
    <link>https://agenticindex.io/rankings/voice-agents</link>
    <atom:link href="https://agenticindex.io/feeds/voice-agents.xml" rel="self" type="application/rss+xml" />
    <description>Verified changes to voice agents, from the Agentic Index change log. Every entry is dated and sourced from public vendor material.</description>
    <language>en-us</language>
    <copyright>Free to reuse with a visible link to agenticindex.io. Terms: https://agenticindex.io/use-this-data</copyright>
    <lastBuildDate>Fri, 25 Sep 2026 12:00:00 GMT</lastBuildDate>
    <ttl>720</ttl>
    <image>
      <url>https://media.base44.com/images/public/69e4cec10b4126d233ee7ee4/a44c0d34a_generated_image.png</url>
      <title>Agentic Index: voice agents change log</title>
      <link>https://agenticindex.io/rankings/voice-agents</link>
    </image>
    <item>
      <title>careCycle expanded its assistant's plan data to every US county across Medicare Advantage and PDP plan types.</title>
      <link>https://agenticindex.io/change-log/carecycle</link>
      <guid isPermaLink="false">agenticindex.io:change:carecycle-1f7by8-2026-09-25</guid>
      <pubDate>Fri, 25 Sep 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Voice agent</category>
      <category>Healthcare agent</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>careCycle expanded its assistant's plan data to every US county across Medicare Advantage and PDP plan types. Its rebuilt lookup pipeline delivers benefit answers in two seconds, down from 6.5 seconds, according to the release note. The same release lets Pulse routing use custom fields and customer assignment use customer tags.</p><p><strong>Why it matters:</strong> Agencies serving multiple regions can assess the assistant against a broader range of member plans and reduce pauses during benefit conversations. Custom routing criteria also let teams reflect their own service assignments without relying only on standard customer attributes.</p><p>Impact: Medium. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://carecycle.ai/changelog/2026-09-25-faster-plan-data">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/carecycle">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Voiceflow made live agent handoff available on every plan.</title>
      <link>https://agenticindex.io/change-log/voiceflow</link>
      <guid isPermaLink="false">agenticindex.io:change:voiceflow-130hl0-2026-09-24</guid>
      <pubDate>Thu, 24 Sep 2026 12:00:00 GMT</pubDate>
      <category>Pricing / packaging</category>
      <category>Customer support agent</category>
      <category>Agent builder</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Voiceflow made live agent handoff available on every plan. Chat agents can use the Handoff step to route conversations to Ujet, Genesys, Kustomer, Dixa, Sunshine Conversations for Zendesk, or a custom handoff provider.</p><p><strong>Why it matters:</strong> Teams on lower plans can include human escalation in a chat deployment without changing their Voiceflow plan to unlock that step. Buyers can assess how the handoff fits their existing contact center and support routing.</p><p>Impact: High. Verification: Partially Verified. Type: Pricing / packaging.</p><p>Evidence: <a href="https://www.voiceflow.com/docs/changelog/changelog">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/voiceflow">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Cartesia introduced Multilingual Voices, a library of more than 50 voices that support up to 25 languages while retaining the same vocal…</title>
      <link>https://agenticindex.io/change-log/cartesia</link>
      <guid isPermaLink="false">agenticindex.io:change:cartesia-pcsges-2026-09-23</guid>
      <pubDate>Wed, 23 Sep 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Agent infrastructure</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Cartesia introduced Multilingual Voices, a library of more than 50 voices that support up to 25 languages while retaining the same vocal identity. Customers can keep a voice ID across its supported languages and select a locale in speech generation requests. The offering is available through the voice library and API, with language and accent support varying by voice.</p><p><strong>Why it matters:</strong> Voice agent teams can expand language coverage without assigning a different speaker identity to each market. Buyers should validate each voice's supported locales and fallback behavior before deploying it across languages.</p><p>Impact: Medium. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://www.cartesia.ai/blog/multilingual-voices">Official blog</a></p><p>Source: <a href="https://agenticindex.io/change-log/cartesia">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Smallest AI added POST /waves/v1/auth/token so servers can issue temporary credentials for browser and mobile speech clients while…</title>
      <link>https://agenticindex.io/change-log/smallest-ai</link>
      <guid isPermaLink="false">agenticindex.io:change:smallest-ai-1gpsnt-2026-09-22</guid>
      <pubDate>Tue, 22 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Voice agent</category>
      <category>Agent infrastructure</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>Smallest AI added POST /waves/v1/auth/token so servers can issue temporary credentials for browser and mobile speech clients while retaining the API key. Tokens last 30 to 900 seconds, defaulting to 300, and authorize supported speech synthesis, transcription, speech to speech, and voice listing routes. They cannot authorize voice cloning, LLM chat completions, analytics, or further token creation; usage remains billed to the issuing key.</p><p><strong>Why it matters:</strong> Teams can expose speech inference to clients with narrower credentials instead of distributing a persistent API key. Integrators must use the same region for token issuance and inference, and account for open WebSocket sessions continuing after token expiry.</p><p>Impact: Medium. Verification: Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://docs.smallest.ai/models/changelog/general/2026/9/22">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/smallest-ai">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>LivePerson changed Connect to Messaging eligibility responses to HTTP 400 when required template variables are missing and no channel…</title>
      <link>https://agenticindex.io/change-log/liveperson</link>
      <guid isPermaLink="false">agenticindex.io:change:liveperson-2nt2fg-2026-09-21</guid>
      <pubDate>Mon, 21 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>LivePerson changed Connect to Messaging eligibility responses to HTTP 400 when required template variables are missing and no channel passes validation, replacing HTTP 200 with eligible set to false. Extra variables remain unaffected. The same release adds agent and simulation assessments to interrupted Syntrix v1.8.1 runs, using only completed conversations. Account delivery dates may vary.</p><p><strong>Why it matters:</strong> Integration owners should test error handling and template payload completeness because unsuccessful handoffs now change HTTP status. Evaluation teams can recover aggregate insights from interrupted simulations without treating incomplete conversations as completed results.</p><p>Impact: Medium. Verification: Partially Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://community.liveperson.com/kb/articles/2382-week-of-sep-21st-2026">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/liveperson">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>ElevenAgents added enable_parallel_tool_calls, enabled by default, so supported models can run multiple tools in one turn.</title>
      <link>https://agenticindex.io/change-log/elevenlabs</link>
      <guid isPermaLink="false">agenticindex.io:change:elevenlabs-11bsc3-2026-09-21</guid>
      <pubDate>Mon, 21 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>ElevenAgents added enable_parallel_tool_calls, enabled by default, so supported models can run multiple tools in one turn. The release also added Slack alerts and a paginated phone number search endpoint. Conversation filters no longer accept the missing operator and instead use an empty value; the knowledge base crawl max_depth parameter is deprecated and ignored.</p><p><strong>Why it matters:</strong> Review tools whose side effects depend on execution order before allowing concurrent calls. Existing integrations must update missing value filters and should no longer rely on crawl depth as a boundary.</p><p>Impact: High. Verification: Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://elevenlabs.io/docs/changelog/2026/9/21">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/elevenlabs">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Vapi added a serviceTier setting that lets assistants request faster OpenAI processing for supported GPT-5.6 and GPT-5.5 models.</title>
      <link>https://agenticindex.io/change-log/vapi</link>
      <guid isPermaLink="false">agenticindex.io:change:vapi-14pxd8-2026-09-21</guid>
      <pubDate>Mon, 21 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Agent infrastructure</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Vapi added a serviceTier setting that lets assistants request faster OpenAI processing for supported GPT-5.6 and GPT-5.5 models. Its September 21 update also adds provider request IDs to call logs for language model, transcription, and voice services when customers bring their own keys.</p><p><strong>Why it matters:</strong> Voice teams can test a latency and cost tradeoff without changing their agent workflow, while request IDs make provider support investigations easier to correlate with specific calls. Fast processing uses the provider's corresponding rates and may be downgraded under load.</p><p>Impact: Medium. Verification: Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://docs.vapi.ai/whats-new/2026/9/21">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/vapi">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Delight.ai introduces Intelligent Voice Handoff, which keeps a caller connected while the agent reaches and briefs the person taking over.</title>
      <link>https://agenticindex.io/change-log/sendbird</link>
      <guid isPermaLink="false">agenticindex.io:change:sendbird-x0rpe6-2026-09-21</guid>
      <pubDate>Mon, 21 Sep 2026 12:00:00 GMT</pubDate>
      <category>Workflow orchestration</category>
      <category>Customer support agent</category>
      <category>Agent infrastructure</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Delight.ai introduces Intelligent Voice Handoff, which keeps a caller connected while the agent reaches and briefs the person taking over. Answer detection distinguishes people from voicemail, while recovery handles busy lines and unanswered calls by returning to the caller, leaving a message, or creating a follow up case. Supported integrations can attach the summary, transcript, and sentiment to a ticket before the transfer.</p><p><strong>Why it matters:</strong> Support teams can preserve the work already completed by an AI agent when a case needs human judgment. Failed transfers retain a follow up path, reducing the risk of callers being stranded between automated and human support.</p><p>Impact: Medium. Verification: Verified. Type: Workflow orchestration.</p><p>Evidence: <a href="https://delight.ai/blog/product/introducing-intelligent-voice-handoff">Official blog</a></p><p>Source: <a href="https://agenticindex.io/change-log/sendbird">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Consio now creates a Re:amaze conversation for each completed inbound or outbound call.</title>
      <link>https://agenticindex.io/change-log/consio</link>
      <guid isPermaLink="false">agenticindex.io:change:consio-1sxt87-2026-09-20</guid>
      <pubDate>Sun, 20 Sep 2026 12:00:00 GMT</pubDate>
      <category>Integrations</category>
      <category>Voice agent</category>
      <category>Customer support agent</category>
      <category>GTM / revenue agent</category>
      <description><![CDATA[<p>Consio now creates a Re:amaze conversation for each completed inbound or outbound call. It links matching customer profiles and adds call details, dispositions, agent notes, AI summaries, and next steps, with a link to the full Consio call record.</p><p><strong>Why it matters:</strong> Support teams using Re:amaze can retain voice interaction context in their existing support workflow and reduce manual call logging.</p><p>Impact: Medium. Verification: Verified. Type: Integrations.</p><p>Evidence: <a href="https://www.consio.ai/updates/reamaze-consio-integration">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/consio">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Assort Health expanded its workflow automation for practices on NextGen Enterprise.</title>
      <link>https://agenticindex.io/change-log/assort-health</link>
      <guid isPermaLink="false">agenticindex.io:change:assort-health-wrv8lr-2026-09-17</guid>
      <pubDate>Thu, 17 Sep 2026 12:00:00 GMT</pubDate>
      <category>Integrations</category>
      <category>Healthcare agent</category>
      <category>Voice agent</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>Assort Health expanded its workflow automation for practices on NextGen Enterprise. Its AI agents can now complete patient access actions directly in the EHR, such as reading open appointment slots and booking them.</p><p><strong>Why it matters:</strong> Write access to the EHR is what lets an agent finish a task instead of handing it to staff. Practices should check how each action appears in the NextGen audit trail.</p><p>Impact: Medium. Verification: Verified. Type: Integrations.</p><p>Evidence: <a href="https://www.prnewswire.com/news-releases/assort-health-expands-direct-ehr-workflow-automation-for-practices-using-nextgen-enterprise-ehr-302881252.html">Press release</a></p><p>Source: <a href="https://agenticindex.io/change-log/assort-health">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Deepgram released Nova-3 Pharma, a speech to text model trained on pharmaceutical vocabulary for accurate drug name recognition in…</title>
      <link>https://agenticindex.io/change-log/deepgram</link>
      <guid isPermaLink="false">agenticindex.io:change:deepgram-d42jtq-2026-09-17</guid>
      <pubDate>Thu, 17 Sep 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Agent infrastructure</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Deepgram released Nova-3 Pharma, a speech to text model trained on pharmaceutical vocabulary for accurate drug name recognition in healthcare voice agents.</p><p><strong>Why it matters:</strong> A misheard drug name is a patient safety error. Domain tuned recognition matters for pharmacy and clinical voice agents, and it should be tested on the buyer's own formulary.</p><p>Impact: Medium. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://developers.deepgram.com/changelog/2026/9/17">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/deepgram">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Zo (Zocdoc): Zocdoc introduced an MCP server as part of its new Care Access Network, letting partners connect to its search, availability and booking…</title>
      <link>https://agenticindex.io/change-log/zocdoc-zo</link>
      <guid isPermaLink="false">agenticindex.io:change:zocdoc-zo-e8ou5e-2026-09-16</guid>
      <pubDate>Wed, 16 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Voice agent</category>
      <category>Healthcare agent</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>Zocdoc introduced an MCP server as part of its new Care Access Network, letting partners connect to its search, availability and booking infrastructure.</p><p><strong>Why it matters:</strong> Real time appointment booking through MCP means AI assistants can book care directly. Health systems should ask how patient data and consent are handled when an assistant books on a patient's behalf.</p><p>Impact: High. Verification: Partially Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://www.prnewswire.com/news-releases/zocdoc-enters-its-next-chapter-expanding-beyond-its-marketplace-to-power-access-to-care-everywhere-302879956.html">Press release</a></p><p>Source: <a href="https://agenticindex.io/change-log/zocdoc-zo">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Consio's agents can now send tracking links by SMS during a call, and the update adds faster product search and improved call routing.</title>
      <link>https://agenticindex.io/change-log/consio</link>
      <guid isPermaLink="false">agenticindex.io:change:consio-1ex35ur-2026-09-15</guid>
      <pubDate>Tue, 15 Sep 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Voice agent</category>
      <category>Customer support agent</category>
      <category>GTM / revenue agent</category>
      <description><![CDATA[<p>Consio's agents can now send tracking links by SMS during a call, and the update adds faster product search and improved call routing.</p><p><strong>Why it matters:</strong> Sending a link mid call closes the loop a voice agent otherwise leaves open. It also means the agent is texting customers, so consent rules for SMS apply.</p><p>Impact: Medium. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://www.consio.ai/updates/ai-sms-call-transfers-multi-currency-transcriptions">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/consio">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Deepgram updated its Python, JavaScript and Java SDKs with new Voice Agent controls, including ForceEndTurn and Flux text to speech…</title>
      <link>https://agenticindex.io/change-log/deepgram</link>
      <guid isPermaLink="false">agenticindex.io:change:deepgram-17ebjco-2026-09-14</guid>
      <pubDate>Mon, 14 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Agent infrastructure</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Deepgram updated its Python, JavaScript and Java SDKs with new Voice Agent controls, including ForceEndTurn and Flux text to speech expressivity controls.</p><p><strong>Why it matters:</strong> ForceEndTurn gives developers a direct way to end an agent's turn, which matters for interruption handling, the hardest part of natural voice conversations.</p><p>Impact: Medium. Verification: Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://developers.deepgram.com/changelog/2026/9/14">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/deepgram">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Smallest AI shipped new MCP tools for latency summaries and prompt cache hit rates, reworked telephony so SIP trunks are managed as…</title>
      <link>https://agenticindex.io/change-log/smallest-ai</link>
      <guid isPermaLink="false">agenticindex.io:change:smallest-ai-1ixp50p-2026-09-14</guid>
      <pubDate>Mon, 14 Sep 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Voice agent</category>
      <category>Agent infrastructure</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>Smallest AI shipped new MCP tools for latency summaries and prompt cache hit rates, reworked telephony so SIP trunks are managed as resources, and released SDK 5.5.0.</p><p><strong>Why it matters:</strong> Latency is the metric voice agent buyers care about most. Exposing it through MCP lets teams monitor and tune it from the tools they already use.</p><p>Impact: Medium. Verification: Partially Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://docs.smallest.ai/voice-agents/changelog/changelog">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/smallest-ai">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Vapi's Composer can now generate simulation scenarios, tester personalities and full test suites automatically, and assistants can fall…</title>
      <link>https://agenticindex.io/change-log/vapi</link>
      <guid isPermaLink="false">agenticindex.io:change:vapi-bulbbo-2026-09-14</guid>
      <pubDate>Mon, 14 Sep 2026 12:00:00 GMT</pubDate>
      <category>Observability / auditability</category>
      <category>Agent infrastructure</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Vapi's Composer can now generate simulation scenarios, tester personalities and full test suites automatically, and assistants can fall back to another Anthropic model on Bedrock from the same provider.</p><p><strong>Why it matters:</strong> Automated test generation lowers the effort of testing a voice agent before launch. Same provider fallback keeps a call alive during a model outage without changing data residency.</p><p>Impact: Medium. Verification: Verified. Type: Observability / auditability.</p><p>Evidence: <a href="https://docs.vapi.ai/whats-new/2026/9/14">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/vapi">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>PolyAI introduced Wren, a new coding agent designed to autonomously analyze production conversations and recommend improvements to dialog…</title>
      <link>https://agenticindex.io/change-log/polyai</link>
      <guid isPermaLink="false">agenticindex.io:change:polyai-k5vsxi-2026-09-13</guid>
      <pubDate>Sun, 13 Sep 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>PolyAI introduced Wren, a new coding agent designed to autonomously analyze production conversations and recommend improvements to dialog agents. Wren operates in the background using a six-step agentic loop to identify issues, test potential fixes, and present actionable recommendations for users to approve or dismiss.</p><p><strong>Why it matters:</strong> This allows business users and operations teams to continuously optimize their AI agents without requiring developer intervention. By automating the evaluation and iteration process, organizations can steadily improve resolution rates and ROI based on live conversation data.</p><p>Impact: High. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://poly.ai/blog/wren-agent-improves-dialog-agents">Official blog</a></p><p>Source: <a href="https://agenticindex.io/change-log/polyai">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Thread announced the general availability of Voice AI, introducing a new method for handling phone-based support.</title>
      <link>https://agenticindex.io/change-log/thread</link>
      <guid isPermaLink="false">agenticindex.io:change:thread-19raoyu-2026-09-13</guid>
      <pubDate>Sun, 13 Sep 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Thread announced the general availability of Voice AI, introducing a new method for handling phone-based support. The feature brings the platform's speed and automation capabilities to voice channels.</p><p><strong>Why it matters:</strong> Support teams can now leverage AI to manage and automate inbound phone calls, reducing manual triage and expanding omnichannel coverage within a single platform.</p><p>Impact: High. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://www.getthread.com/service-magic-blog/thread-officially-launches-voice-ai">Official blog</a></p><p>Source: <a href="https://agenticindex.io/change-log/thread">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Smallest AI released an integration with Pipecat, the open source framework for voice and multimodal agents, so developers can use…</title>
      <link>https://agenticindex.io/change-log/smallest-ai</link>
      <guid isPermaLink="false">agenticindex.io:change:smallest-ai-17e9lmr-2026-09-12</guid>
      <pubDate>Sat, 12 Sep 2026 12:00:00 GMT</pubDate>
      <category>Integrations</category>
      <category>Voice agent</category>
      <category>Agent infrastructure</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>Smallest AI released an integration with Pipecat, the open source framework for voice and multimodal agents, so developers can use Smallest AI's voice stack inside Pipecat.</p><p><strong>Why it matters:</strong> Pipecat is where many teams assemble voice agents from parts. Native support lowers the switching cost to try Smallest AI's models without rewriting the pipeline.</p><p>Impact: Medium. Verification: Verified. Type: Integrations.</p><p>Evidence: <a href="https://smallest.ai/blog/smallest-ai-pipecat-the-voice-stack-you-ve-been-waiting-to-build-on">Official blog</a></p><p>Source: <a href="https://agenticindex.io/change-log/smallest-ai">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Outcraft AI introduced a new per-lead pricing model for its AI SDR platform, replacing traditional per-seat and per-minute billing.</title>
      <link>https://agenticindex.io/change-log/outcraft-ai</link>
      <guid isPermaLink="false">agenticindex.io:change:outcraft-ai-180c0cn-2026-09-10</guid>
      <pubDate>Thu, 10 Sep 2026 12:00:00 GMT</pubDate>
      <category>Pricing / packaging</category>
      <category>GTM / revenue agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Outcraft AI introduced a new per-lead pricing model for its AI SDR platform, replacing traditional per-seat and per-minute billing. The new structure charges a flat monthly rate based on the number of inbound leads the AI actively engages, starting at $300 per month for up to 100 leads.</p><p><strong>Why it matters:</strong> Sales teams can now align their software costs directly with actual agent workload and lead volume rather than headcount or call duration. The inclusion of unlimited seats and all communication channels in every plan also removes friction for scaling revenue teams.</p><p>Impact: High. Verification: Partially Verified. Type: Pricing / packaging.</p><p>Evidence: <a href="https://www.globenewswire.com/news-release/2026/09/10/3359909/0/en/ai-sdr-startup-outcraft-ai-rolls-out-per-lead-pricing-for-its-inbound-sales-agents.html">Press release</a></p><p>Source: <a href="https://agenticindex.io/change-log/outcraft-ai">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>SoundHound AI has officially completed its acquisition of LivePerson.</title>
      <link>https://agenticindex.io/change-log/liveperson</link>
      <guid isPermaLink="false">agenticindex.io:change:liveperson-ewb0b9-2026-09-04</guid>
      <pubDate>Fri, 04 Sep 2026 12:00:00 GMT</pubDate>
      <category>Funding / partnership</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>SoundHound AI has officially completed its acquisition of LivePerson. The integration combines SoundHound's voice AI capabilities with LivePerson's digital chat and messaging platform into a unified omnichannel solution.</p><p><strong>Why it matters:</strong> Enterprise buyers can now leverage a single conversational AI platform for both voice and digital messaging touchpoints. The acquisition also brings financial stability and expanded agentic AI capabilities under the SoundHound AI umbrella.</p><p>Impact: High. Verification: Partially Verified. Type: Funding / partnership.</p><p>Evidence: <a href="https://www.globenewswire.com/news-release/2026/09/04/3356596/0/en/soundhound-ai-completes-acquisition-of-liveperson-creating-a-world-leading-omnichannel-conversational-ai-powerhouse.html">Press release</a></p><p>Source: <a href="https://agenticindex.io/change-log/liveperson">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>SoundHound AI has partnered with Perplexity to integrate real-time web search capabilities into SoundHound Chat AI.</title>
      <link>https://agenticindex.io/change-log/soundhound</link>
      <guid isPermaLink="false">agenticindex.io:change:soundhound-1mj3cpp-2026-09-03</guid>
      <pubDate>Thu, 03 Sep 2026 12:00:00 GMT</pubDate>
      <category>Integrations</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>SoundHound AI has partnered with Perplexity to integrate real-time web search capabilities into SoundHound Chat AI. The voice assistant can now respond to user questions conversationally using up-to-date knowledge retrieved directly from the web.</p><p><strong>Why it matters:</strong> Buyers deploying SoundHound Chat AI can offer end users more accurate and timely answers to dynamic queries. This integration reduces hallucinations and expands the assistant's utility beyond static training data.</p><p>Impact: Medium. Verification: Verified. Type: Integrations.</p><p>Evidence: <a href="https://www.soundhound.com/newsroom/soundhound-ai-and-perplexity-join-forces-to-supercharge-voice-assistants">Official announcement</a></p><p>Source: <a href="https://agenticindex.io/change-log/soundhound">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>LivePerson announced updated cutover dates for migrating aiStudio custom flows to Google Gemini models.</title>
      <link>https://agenticindex.io/change-log/liveperson</link>
      <guid isPermaLink="false">agenticindex.io:change:liveperson-1ulzmay-2026-09-02</guid>
      <pubDate>Wed, 02 Sep 2026 12:00:00 GMT</pubDate>
      <category>Deployment / data residency</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>LivePerson announced updated cutover dates for migrating aiStudio custom flows to Google Gemini models. The cutover for all AI agents will occur on September 3, 2026, for EMEA brands and September 17, 2026, for North America brands.</p><p><strong>Why it matters:</strong> Administrators must observe the designated freeze periods for creating or modifying flows to ensure a smooth transition to the new Gemini models. This migration changes the underlying generative AI infrastructure to enhance agent performance.</p><p>Impact: High. Verification: Verified. Type: Deployment / data residency.</p><p>Evidence: <a href="https://community.liveperson.com/kb/articles/2379-week-of-aug-26th-2026">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/liveperson">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Vapi introduced native outbound calling campaigns, allowing users to upload contacts, select an assistant, and track results directly from…</title>
      <link>https://agenticindex.io/change-log/vapi</link>
      <guid isPermaLink="false">agenticindex.io:change:vapi-rf48l9-2026-08-31</guid>
      <pubDate>Mon, 31 Aug 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Agent infrastructure</category>
      <category>Customer support agent</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Vapi introduced native outbound calling campaigns, allowing users to upload contacts, select an assistant, and track results directly from the dashboard. The update also adds a prompt editor with versioning snapshots, a new webhook for processing call artifacts as they upload, and Azure region pinning for GPT-5 models.</p><p><strong>Why it matters:</strong> Native campaign management eliminates the need for third-party dialing orchestration, simplifying outbound operations. Azure region pinning also allows enterprise buyers to enforce data residency and reduce latency, while the new prompt editor and versioning tools improve the developer experience for maintaining agent configurations.</p><p>Impact: High. Verification: Partially Verified. Type: Agent capability.</p><p>Evidence: <a href="https://docs.vapi.ai/whats-new">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/vapi">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Secrets and third-party integrations can now be managed programmatically via the stable API, CLI, and the Voiceflow MCP server.</title>
      <link>https://agenticindex.io/change-log/voiceflow</link>
      <guid isPermaLink="false">agenticindex.io:change:voiceflow-cin2or-2026-08-31</guid>
      <pubDate>Mon, 31 Aug 2026 12:00:00 GMT</pubDate>
      <category>MCP / tool calling / API</category>
      <category>Customer support agent</category>
      <category>Agent builder</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Secrets and third-party integrations can now be managed programmatically via the stable API, CLI, and the Voiceflow MCP server. Secret values are strictly write-only and are automatically refused if they arrive from an MCP client.</p><p><strong>Why it matters:</strong> Engineering teams can fully automate the provisioning of agent environments and securely manage credentials in CI/CD pipelines without requiring manual intervention in the visual builder.</p><p>Impact: High. Verification: Partially Verified. Type: MCP / tool calling / API.</p><p>Evidence: <a href="https://www.voiceflow.com/docs/changelog">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/voiceflow">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>careCycle introduced Assembly, a new campaign workspace for configuring AI employees.</title>
      <link>https://agenticindex.io/change-log/carecycle</link>
      <guid isPermaLink="false">agenticindex.io:change:carecycle-15c55hf-2026-08-28</guid>
      <pubDate>Fri, 28 Aug 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Voice agent</category>
      <category>Healthcare agent</category>
      <category>Customer support agent</category>
      <description><![CDATA[<p>careCycle introduced Assembly, a new campaign workspace for configuring AI employees. The feature allows users to query current configurations, review call performance, and prepare changes via a guided conversation, complete with full version control for tracking and restoring previous setups.</p><p><strong>Why it matters:</strong> This provides administrators with greater control and visibility over their AI agents' behavior. The built-in version history ensures teams can safely iterate on campaign configurations and easily roll back if needed.</p><p>Impact: High. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://carecycle.ai/changelog/2026-08-28-assembly">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/carecycle">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Consio introduced a unified call history view across workspaces with CSV export capabilities that include call transcripts.</title>
      <link>https://agenticindex.io/change-log/consio</link>
      <guid isPermaLink="false">agenticindex.io:change:consio-uz2u1d-2026-08-28</guid>
      <pubDate>Fri, 28 Aug 2026 12:00:00 GMT</pubDate>
      <category>Observability / auditability</category>
      <category>Voice agent</category>
      <category>Customer support agent</category>
      <category>GTM / revenue agent</category>
      <description><![CDATA[<p>Consio introduced a unified call history view across workspaces with CSV export capabilities that include call transcripts. The update also allows merchants outside the US and Canada to register phone numbers without encountering blocked trust registration steps.</p><p><strong>Why it matters:</strong> The unified history and CSV exports improve auditing and quality assurance for customer interactions. Additionally, international merchants can now onboard and provision phone numbers more easily.</p><p>Impact: Medium. Verification: Verified. Type: Observability / auditability.</p><p>Evidence: <a href="https://www.consio.ai/updates/call-history-and-csv-export">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/consio">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Gnani.ai launched Artha, a sovereign enterprise AI stack that combines its new 30-billion-parameter open-weight foundational model, Evon…</title>
      <link>https://agenticindex.io/change-log/gnani-ai</link>
      <guid isPermaLink="false">agenticindex.io:change:gnani-ai-10h1jpo-2026-08-28</guid>
      <pubDate>Fri, 28 Aug 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Voice agent</category>
      <category>Customer support agent</category>
      <category>Agent infrastructure</category>
      <description><![CDATA[<p>Gnani.ai launched Artha, a sovereign enterprise AI stack that combines its new 30-billion-parameter open-weight foundational model, Evon v3.3, with the Plexus platform for AI agents.</p><p><strong>Why it matters:</strong> Enterprise buyers can now leverage a localized, sovereign AI stack specifically designed for agentic workflows. This enables organizations to build and deploy highly capable voice agents while maintaining strict data residency and control.</p><p>Impact: High. Verification: Partially Verified. Type: Agent capability.</p><p>Evidence: <a href="https://economictimes.indiatimes.com/ai/ai-insights/gnani-ai-launches-artha-sovereign-ai-stack-with-30-billion-parameter-evon-3-3/articleshow/133592679.cms">News article</a></p><p>Source: <a href="https://agenticindex.io/change-log/gnani-ai">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Cartesia released Sonic 3.6, an updated text-to-speech model that adapts intonation, pacing, and emotiveness to transcript context without…</title>
      <link>https://agenticindex.io/change-log/cartesia</link>
      <guid isPermaLink="false">agenticindex.io:change:cartesia-qmqpcx-2026-08-27</guid>
      <pubDate>Thu, 27 Aug 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Agent infrastructure</category>
      <category>Voice agent</category>
      <description><![CDATA[<p>Cartesia released Sonic 3.6, an updated text-to-speech model that adapts intonation, pacing, and emotiveness to transcript context without SSML tags. The release adds support for Odia and Urdu, bringing the total to 44 supported languages, and improves transcript following for Hindi and Hinglish.</p><p><strong>Why it matters:</strong> Voice agent developers can deliver more expressive conversational experiences without altering existing code or voice IDs. The improved handling of non-English alphanumerics and in-transcript disfluencies reduces the need for complex prompt engineering to achieve natural hesitation.</p><p>Impact: High. Verification: Partially Verified. Type: Agent capability.</p><p>Evidence: <a href="https://docs.cartesia.ai/changelog/2026">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/cartesia">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
    <item>
      <title>Consio AI agents can now create Shopify draft orders directly during a customer call.</title>
      <link>https://agenticindex.io/change-log/consio</link>
      <guid isPermaLink="false">agenticindex.io:change:consio-17en8xe-2026-08-27</guid>
      <pubDate>Thu, 27 Aug 2026 12:00:00 GMT</pubDate>
      <category>Agent capability</category>
      <category>Voice agent</category>
      <category>Customer support agent</category>
      <category>GTM / revenue agent</category>
      <description><![CDATA[<p>Consio AI agents can now create Shopify draft orders directly during a customer call. The agent can select the requested product, apply discount codes, and generate a draft order ready for the customer to complete.</p><p><strong>Why it matters:</strong> This capability transforms the AI agent from a purely support-focused tool into a sales-driving asset. Merchants can now capture high-intent buyers immediately on the phone without requiring human intervention.</p><p>Impact: High. Verification: Verified. Type: Agent capability.</p><p>Evidence: <a href="https://www.consio.ai/updates/your-ai-agent-can-now-close-the-sale">Official changelog</a></p><p>Source: <a href="https://agenticindex.io/change-log/consio">Agentic Index change log</a>. Free to reuse with a link back. <a href="https://agenticindex.io/use-this-data">Terms</a>.</p>]]></description>
    </item>
  </channel>
</rss>
