Freshworks Unveils Enhanced Freddy Agentic AI Platform with Autonomous Problem-Solving Capabilities

Freshworks Upgrades Its AI Platform With Abilities to Create AI Agents and More | AIM

Freshworks says Freddy Agentic AI Platform is designed to go beyond simple automation to solve challenges in real-time. Freshworks has launched new features for its Freddy AI platform, enhancing automation across customer and employee service operations. The updates aim to simplify complex workflows and enable autonomous resolution of service requests. The upgraded platform includes Freddy AI Agent Studio, a no-code builder for deploying task-specific AI agents, and new capabilities in email support, knowledge search, IT insights, and AI copilot functions. These features allow Freddy AI agents to take direct actions, such as issuing refunds or booking appointments across business apps like Shopify and Stripe, without human intervention. "AI agents can be taught unique skills to autonomously take actions like issuing a refund, checking order status, or updating a customer record," the company stated. Businesses also get access to pre-built templates to easily add the required skills to the AI agents they create. Murali Swaminathan, CTO of Freshworks, told AIM, "The whole idea of the Freddie AI agent studio is to provide a set of out-of-the-box templates for our customers to build AI agents quickly. We also have our Skill Builder behind the scenes, which will enable them to create new AI agents." The new Freddy Agentic AI platform supports multi-application workflows, autonomous email handling, and root cause analysis for IT operations. The agent integrates with Slack and Microsoft tools, supports 40+ languages, and offers improved context awareness through a multi-LLM architecture. Swaminathan highlighted that the company utilises GPT-4o and GPT-4o mini models under the hood and hosts several others, including Llama, for specialised applications. The company claims real-world productivity gains for its customers. Swaminathan told AIM, "Hobbycraft is our customer that has deployed Freddy AI agents for the customer service use case. They've been able to successfully deflect at least 30% of their support requests, which has indirectly helped them improve customer satisfaction by at least 25%." The company also mentioned other customer stories, such as Bergzeit, which reduced ticket translation efforts by 75%, and iPostal1, which now resolves over half of its support queries automatically. Furthermore, Five9 reports saving 200 IT hours monthly using Freddy AI Copilot. Swaminathan explained to AIM that the company is exploring opening up an MCP server to Freddy AI capabilities and is aiming for wider collaboration with other leading solutions, like a Salesforce agent. Furthermore, Agent2Agent (A2A) protocols are also being explored to facilitate interoperability between AI agents beyond the Freshworks ecosystem. To encourage adoption, Freshworks has introduced training resources, partner programmes, and professional services. With over 5,000 organisations using Freddy AI since its 2023 launch, Freshworks positions its agentic platform as a scalable alternative to legacy service automation.

Freshworks debuts upgraded Freddy Agentic AI Platform at Refresh event - SiliconANGLE

Freshworks debuts upgraded Freddy Agentic AI Platform at Refresh event Customer service and support software firm Freshworks Inc. today debuted an expanded version of its Freddy Agentic AI Platform at its annual Refresh event, adding new capabilities that go beyond answering service queries to autonomously resolving them. The Freddy AI Agent, launched in October, was designed to improve customer and employee experiences by autonomously resolving a significant portion of service requests. The new, updated version of the Freddy Agentic platform offers a connected, intelligent, continuously learning system of AI agents that don't just reply to service questions, but can resolve them. The new update also includes the introduction of the Freddy AI Agent Studio, a no-code platform that allows customer service teams to build and deploy custom AI agents without needing technical expertise. The agents are capable of independently handling a wide range of tasks, from issuing refunds and updating payroll records to booking shipments and managing loyalty rewards. The idea is to help businesses reduce repetitive workloads, accelerate resolution times and improve overall service quality. According to Freshworks, the new, easy-to-deploy AI agents can think, reason and act to resolve many customer queries from start to finish, giving human agents more time to handle more complex customer challenges. "We're on a mission to uncomplicate the most grueling and repetitive work that IT service and customer support teams face every day," said Chief Executive Dennis Woodside. "Just like our core software, Freddy Agentic AI gets up and running fast and delivers immediate value -- serving as a business accelerator, not another overcomplicated project." The updated Freddy Agentic AI Platform has already been tested among some of Freshworks' customers and, according to the company, is already delivering tangible value across industries including retail, finance, software and manufacturing. Early adopters such as Hobbycraft Trading Ltd., iPostal1 Inc. and Five9 Inc. have reported measurable improvements in efficiency and satisfaction metrics, including up to 70% ticket deflection and significant time savings for support teams, Freshworks says. To accelerate adoption, the company is also offering new training resources, professional services and in-product guidance to help customers scale up AI usage quickly and effectively. "We designed our Freddy Agentic AI Platform to go beyond simple automation. It works alongside people to solve real challenges in real time," said Chief Product Officer Srini Raghavan. "Freddy's multimodel architecture draws on our trusted LLM partners, each selected for their strengths. This layered approach helps ensure more accurate, reliable and context-aware support that helps resource constrained teams move faster and with less friction."