AI agents move from tools to operators
AI agents funding refers to investment flowing into software systems that can observe, decide and execute business workflows autonomously, reducing manual repetition and turning narrow tools into task-owning digital operators. Investors are now backing a new wave of workflow automation startups that promise more than chat interfaces or simple macros. These AI agents watch how people work, infer intent and then perform tasks in the background across applications and data sources. Unlike earlier automation efforts that relied on rigid rules or scripts, agentic platforms aim to support open-ended, evolving processes in business operations automation, from email handling to complex commercial workflows. This shift reframes software as a co-worker that can learn, adapt and coordinate actions over time, which is why venture capital is clustering around companies that position AI agents as the next operating layer for knowledge and operations teams.
IrisGo: a proactive desktop agent for knowledge workers
IrisGo shows how investors are backing AI agents that learn workflows directly from user behavior. The desktop agent observes a person complete a task once—ordering coffee, processing an invoice, drafting an email—then automates the same workflow without repeat instructions. Its skills library covers common knowledge-worker tasks, while an embedded coding assistant brings developer-focused capabilities into the same environment. The company raised USD 2.8 million (approx. RM12.9 million) in seed funding led by Andrew Ng’s AI Fund, with Nvidia and Google also participating, to pursue this proactive agent model. A hybrid on-device and cloud design targets privacy-conscious teams, with cloud processing requiring explicit user approval. Early momentum includes beta releases for macOS and Windows and a preinstallation deal with Acer, signalling how workflow automation startups can reach users directly at the device level rather than only through SaaS channels.
Kopa.ai and the rise of e-commerce AI agents
In e-commerce, Kopa.ai is building AI agents that behave less like isolated tools and more like embedded operators inside online shops. Its platform connects to existing storefronts and tools, then runs a continuous loop of analysis, decision-making and execution across marketing campaigns, inventory, site performance and customer behaviour. Instead of forcing teams to craft prompts or predefined flows, merchants set high-level goals and the system determines which actions to take, from generating creatives to reallocating budgets or publishing updates. According to Kopa.ai, the aim is to handle “thousands of expert decisions every week” that online retailers face as operations scale. The company secured €2 million to expand its agentic infrastructure, betting that e-commerce AI agents can reduce operational drag and give teams a way to delegate recurring work while retaining control through optional human approvals.
Handshake: automating complex retail buying workflows
Specialist AI agents are also emerging for narrow, high-value workflows such as retail buying. Handshake focuses on the messy reality of retail dealmaking, where buyers and suppliers rely on complex agreements scattered across long email trails. Its platform brings these agreements into a single system where commercial teams can make, track and execute deals with more structure and transparency. The company raised USD 3.2 million (approx. RM14.8 million) led by Triple Point Ventures to build what its founders call an “agentic AI operating system for dealmaking”. By embedding AI into the agreement lifecycle, Handshake aims to automate repetitive follow-ups, highlight missed commitments and adapt to the wide variety of deal structures buyers encounter. For investors, this kind of domain-specific automation shows how AI agents can address operational pain points that traditional enterprise software and generic CRMs have struggled to cover.
Why investors are betting on operational automation
Across IrisGo, Kopa.ai and Handshake, a common thesis is pulling capital into AI agents funding: these systems reduce manual task repetition and free teams to focus on judgment, relationships and strategy. Investors see that many organisations already adopted AI in a piecemeal way—chatbots, copilots, basic scripting—but still rely on people to glue everything together. Agentic platforms promise to sit in that glue layer, orchestrating tools, data and processes as a “workflow autopilot” rather than a passive assistant. The appeal is both horizontal and vertical: desktop agents that help any knowledge worker, e-commerce AI agents tuned to online retail, and dealmaking systems tailored to commercial teams. If these startups can prove reliability, governance and measurable productivity gains, AI agents could become a default layer of business operations automation, turning today’s experiments into core infrastructure.