In an era where AI is rapidly transforming workplaces, one company is pioneering a shift from simple chatbot interactions to enabling AI tools as vital, intuitive companions for everyday tasks. But here’s where it gets controversial — many still wonder whether these advanced AI agents truly enhance productivity or just create new layers of complexity. Let’s explore how one innovative startup is redefining the landscape.
Eric Jing, a visionary entrepreneur, is committed to moving beyond conventional AI chatbots to develop tools that serve as seamless extensions of everyday work routines. His goal is for users not simply to converse with AI, but to engage with it as an active partner that delivers tangible business outcomes—all from a single, integrated platform.
This vision is embodied in Genspark’s Super Agent, a product of Mainfunc—an AI startup Jing co-founded with Kay Zhu in Palo Alto in 2024, with operations also spanning Singapore and Japan. Unlike traditional large language models (LLMs) used for writing, coding, or solving problems, Genspark’s approach emphasizes direct interaction with specific goals. Jing explains, “We want our users to communicate directly with the desired output, rather than just the AI itself.” For example, if the goal is to create a presentation slide, the user prompts the agent, and it produces the slide—no need to deal with intermediary steps or formats. The same applies for spreadsheets or other outputs; the user interacts with the final product, not with the underlying AI.
Currently, Genspark boasts over 80 specialized, self-developed agents capable of a wide array of tasks—from editing videos and generating images to building entire websites, designing marketing materials, or even booking restaurant reservations. Such diversity demonstrates how AI can support multiple facets of business and creativity.
And the ecosystem is expanding. As announced today at Microsoft Ignite, agents from Genspark, Cognition, Glean, Kasisto, n8n, and others will now be accessible within Microsoft’s new Agent 365 platform. This availability allows organizations to discover and deploy these agents securely, integrating seamlessly into their existing digital environments.
Jing emphasizes that this partnership aims to embed Genspark into the larger Microsoft ecosystem. “Our goal is to ensure knowledge workers can accomplish tasks not just on our platform but also within Microsoft 365, without needing to switch between different tools,” he says. This integration reduces friction and accelerates productivity, making AI a more natural part of everyday workflows.
Across the globe, companies are adopting thousands of similar AI agents to automate routine tasks—Microsoft’s 2025 Work Trend Index projects this number will surpass 1.3 billion by 2028. But with this explosive growth comes a critical challenge: managing these agents effectively. Who created them? What data can they access? Are they behaving safely and ethically? Without proper oversight, unmanaged agents could become security risks, leak sensitive information, or operate outside policies.
To address this, Microsoft launched Agent 365, a comprehensive control system that extends existing infrastructure to include AI agents. It grants these agents access to the same productivity tools, with built-in protections via Microsoft security and governance solutions like Defender, Entra, and Purview. Additionally, IT teams can oversee agent activities through the Microsoft 365 admin center, ensuring compliance, safety, and collaboration with human managers across various workflows.
Irina Nechaeva, Microsoft’s general manager of identity product marketing, underscores the importance of these controls, stating that managing AI agents must go hand-in-hand with developing their capabilities. “It’s crucial to have the right tools to monitor, protect, and optimize these agents right from the start,” she notes.
Imagine a busy marketing executive tasked with creating a multimedia proposal for the company’s leadership. With Microsoft Agent 365 and Genspark’s Super Agent—a readily available AI assistant—the process becomes streamlined. The user simply adds the agent to their team within the Microsoft ecosystem. The agent, equipped with a secure identifier, connects to the necessary resources. It then researches, generates slides, documents, videos, and posters, refining these materials based on ongoing discussions via email, chat, or Teams meetings. When the project concludes, the system verifies the agent’s activities against company policies, ensuring accountability.
Jing is optimistic about the impact of integrating Genspark’s unique capabilities into Microsoft’s platform. He believes it will create “magic moments” for users who seek intuitive, powerful tools. For instance, Genspark’s slide creation tool is already optimized for PowerPoint, allowing users to export completed slides directly into the familiar presentation format—meaning users can stick with the tools they already trust.
However, Jing is quick to highlight that human oversight remains vital. “AI can handle 80% to 90% of the preliminary work,” he explains, “but the final 10% still requires human judgment and verification.” This balance ensures quality while leveraging AI’s efficiencies.
Partnering with Microsoft is a significant milestone for Mainfunc, a relatively new player founded just a year ago, with a modest team of 30 dedicated professionals. Focused on helping knowledge workers do more, Mainfunc’s single-minded approach has earned it $50 million in annual recurring revenue and recognition as one of the fastest-growing AI startups worldwide.
Genspark’s technical backbone combines eight leading LLMs, including OpenAI’s GPT and Anthropic’s Claude, providing tailored, purpose-specific services supported by Microsoft Azure’s robust cloud infrastructure. Jing, a former member of Microsoft’s Bing search team, emphasizes that this focus on user-centered solutions and reliable cloud hosting is key to their success.
Azure’s flexibility allows Genspark to interface with multiple AI models from providers like OpenAI and Anthropic through a unified platform, streamlining deployment and management.
Looking ahead, Jing believes that the AI revolution is just beginning. With current adoption levels less than 1%, he sees enormous potential for growth. “In two years, we’ll see a shift from developer-focused AI into mainstream usage among knowledge workers,” he predicts, confident that AI will bring substantial improvements to work quality and efficiency.
So, the question remains: Are we truly ready for an AI-powered future where strategic oversight is just as critical as technological innovation? And what happens when AI agents operate beyond our control? Share your thoughts—do you see AI as a chance to elevate human effort or as a risk that needs careful management?