Imagine a world where AI isn't just a chatbot, but a true partner that understands your goals and helps you achieve them. That's the bold vision driving a new wave of innovation, and Microsoft's Agent 365 is at the forefront. But how do we ensure these powerful AI assistants are secure and effective? Let's dive in.
Eric Jing, a visionary entrepreneur, is determined to revolutionize how we interact with AI. His mission? To seamlessly integrate AI into our daily workflows, transforming it from a simple chatbot into an indispensable companion. Jing envisions a future where AI empowers us to achieve our goals effortlessly, all within a unified platform.
That platform is Genspark Super Agent, the brainchild of Mainfunc, an AI startup Jing co-founded with Kay Zhu in Palo Alto, California, back in 2024. Mainfunc has since expanded its reach with offices in Singapore and Japan, demonstrating its global ambitions.
"We want our users to talk directly to the business outcome, instead of talking to a chatbot. So that’s quite a different user experience," Jing explains. He distinguishes Genspark from traditional Large Language Models (LLMs), the generative AI engines commonly used for writing, coding, and problem-solving. Genspark offers a radically different user experience.
On Genspark, users interact with digital agents using natural language, prompting them to create slides, spreadsheets, and various other deliverables directly. Think of it as having a conversation with the final product itself.
"If the outcome is a slide, you are talking to the slide directly. If the outcome you want is a spreadsheet, you talk to the spreadsheet," Jing elaborates. This direct interaction streamlines the process and focuses on the desired result.
Genspark boasts over 80 homegrown agents, each designed to deliver specific services. These agents can create edited videos, generate stunning images, build entire websites, craft compelling marketing materials, and even make restaurant reservations. The possibilities seem endless!
And now, in a significant development, agents from Genspark and other leading providers, including Cognition, Glean, Kasisto, and n8n, will be securely accessible to organizations through Microsoft's Agent 365, unveiled recently at Ignite. This integration marks a major step forward in the adoption of AI agents in the enterprise.
"We want to make Genspark… a part of the Microsoft ecosystem," Jing states, highlighting the strategic importance of partnering with Microsoft. "So a knowledge worker can get tasks done not just from the Genspark site, but also within the Microsoft 365 platform without leaving it." This seamless integration promises to boost productivity and efficiency for users.
According to Microsoft's 2025 Work Trend Index report, firms worldwide are already deploying thousands of agents to automate routine tasks. And the trend is only accelerating. An IDC Info Snapshot, sponsored by Microsoft, predicts that the number of deployed agents will skyrocket to over 1.3 billion by 2028. This explosive growth underscores the transformative potential of AI agents in the workplace.
But here's where it gets controversial... With such a massive influx of AI agents, managing them effectively becomes paramount. Keeping track of who created these agents, what data they can access, and whether they're behaving ethically and securely is absolutely crucial. Unmanaged agents can quickly become a liability, posing a significant threat to data security and compliance.
Agent 365 offers a solution: a centralized control plane for managing AI agents. It extends existing infrastructure used for managing human employees to encompass AI agents, equipping them with the same access controls and security protocols. This streamlined approach reduces the time and effort required to integrate agents into existing business processes.
Agent 365 leverages Microsoft's robust security and productivity solutions, including Defender, Entra, and Purview, to protect and govern AI agents. IT departments can manage agents through the familiar Microsoft 365 admin center, enabling seamless collaboration between AI agents and their human counterparts across Microsoft 365 apps and other workflows.
Irina Nechaeva, general manager of identity product marketing at Microsoft, emphasizes the importance of proactive security measures. "All of this needs to be happening together with the agentic innovation itself. So not just the agents, but also the right tools to control, monitor, protect, and make those agents more effective." In other words, security must be built in from the ground up, not added as an afterthought.
So, what does this all mean in practice? Imagine a time-strapped marketing executive who needs to create a multimedia proposal for the board. How can Agent 365 and Genspark streamline this process?
The executive could access Microsoft Agent 365, select Genspark's Super Agent, and seamlessly add it to their team. Thanks to the agent's built-in Microsoft Entra Agent ID, it can securely access the necessary resources to complete the task. The executive could then prompt the agent to research the topic, create a presentation deck, draft a document, design a website, produce a video, and even generate eye-catching posters. The agent can then refine and update these materials based on feedback received through email, chat, or Teams meetings.
At the end of the project, Agent 365 can even compare the agent's logged activities against the firm's guidelines and policies, ensuring compliance and accountability. This level of transparency and control is essential for building trust in AI agents.
Jing expresses confidence that integrating Genspark's "unique capabilities" into the Microsoft ecosystem through Agent 365 will create "magic moments" for users. He points out that Genspark's AI tool for creating slides was specifically designed with Microsoft PowerPoint in mind.
"We wrote four iterations of the file exporting feature so you can export a Genspark slide into a perfect PowerPoint file… people already familiar with PowerPoint or with this kind of workflow in their company—they can still stick to PowerPoint," he explains. This seamless integration with existing tools and workflows minimizes disruption and maximizes user adoption.
And this is the part most people miss... Jing stresses that even with the advancements in AI, the human element remains crucial.
"In our world, we believe that 80% to 90% of the preliminary work can be done by AI. But the last mile – 5% to 10% of work – should still be done by humans manually. You need to verify everything," he emphasizes. This hybrid approach, combining the efficiency of AI with the critical thinking and judgment of humans, is the key to success.
Partnering with Microsoft represents a significant milestone for Mainfunc, a one-year-old startup, and its multi-agent platform, Genspark, which launched in April 2025.
With a lean team of just 30 people, Mainfunc is laser-focused on one goal: helping knowledge workers achieve more. Jing sees Microsoft as providing a "great channel to enterprise," a user segment that Mainfunc has limited access to on its own. "And it’s a great marketplace for different software," he adds.
By concentrating its efforts on identifying the AI models that can best address the pain points of knowledge workers at the lowest cost, Mainfunc has achieved impressive results. The company boasts an annual recurring revenue (ARR) of $50 million and has been recognized as one of the world's fastest-growing AI startups.
Genspark's agents are powered by a carefully curated mix of eight LLMs, including industry giants like OpenAI's GPT and Anthropic's Claude. This blend allows Genspark to deliver a best-for-purpose service, underpinned by the robust infrastructure of Microsoft Azure.
"This is definitely our secret sauce," says Jing, an industry veteran who was a founding team member for Microsoft's search engine, Bing. His deep understanding of both AI technology and user needs has been instrumental in Mainfunc's success.
This unwavering focus on user solutions, coupled with Jing's prior experience at Microsoft, led Mainfunc to adopt Microsoft Azure as a reliable and secure infrastructure for Genspark.
"Basically, I want to… solve the pain point from our users, not solve the pain point of different services," Jing explains. "We trust Azure to host scalable cloud solutions." Azure's scalability and security are essential for supporting Genspark's growing user base and complex AI models.
Azure also enables Genspark to seamlessly interface with multiple AI models, including those from OpenAI and Anthropic, through a single platform, simplifying development and deployment.
According to Jing, Mainfunc is on track based on AI startups' "true north" metrics, which measure survival and growth. In addition to its impressive ARR, Genspark achieved a paid retention rate of 88% to 92% in its first month, demonstrating strong customer loyalty. Furthermore, its monthly burn rate averaged less than $1 million for the second quarter ended June 30, 2025, indicating efficient financial management.
Jing identifies Japan, the US, and South Korea as Mainfunc's top subscriber markets.
He foresees tremendous growth potential for the AI sector, believing that the AI revolution has only penetrated "less than 1% of society." He emphasizes that there is still significant room for improvement in the quality of work that AI can deliver.
"Two years from now, I see the entire market will shift more and more from developer communities to the knowledge workers," he predicts. This shift will require a new generation of AI tools and platforms designed specifically for the needs of knowledge workers.
Now, let's open the floor for discussion. Do you agree with Jing's vision of AI transforming the workplace? How concerned are you about the security risks associated with unmanaged AI agents? And what role do you see for human oversight in the age of AI? Share your thoughts in the comments below!
