From Chatbots to Colleagues: The New Wave of Enterprise AI Agents
AI agents in enterprise are software systems that can understand goals, act independently across business applications, complete end-to-end tasks such as accounting entries or customer responses, and coordinate with human teams, turning automation from tool-based assistance into workflow-level execution.
The latest burst of enterprise AI funding is not chasing another generic chatbot. It is backing AI agents that behave like specialized digital employees embedded in specific workflows. Finto raised USD 3.4 million (approx. RM15.6 million) to place AI agents inside enterprise accounting, automating invoice handling instead of only improving OCR or templates. AIsa secured USD 6.5 million (approx. RM29.9 million) to build a transaction network so AI agents can discover and pay for models, APIs, data, and SaaS tools without human billing work. Sable pulled in USD 45 million (approx. RM207 million) to scale Aidan, an AI “employee” that conducts customer interactions through live product demonstrations and software control. Meanwhile, Medallia’s recapitalization locks in USD 500 million (approx. RM2.3 billion) of product investment aimed at agentic workflows in customer experience platforms.
Accounting Gets Its First True AI Colleague
Enterprise accounting has been one of the most stubbornly manual corners of back-office work, and that is precisely why AI automation accounting is attracting targeted capital. Finto’s USD 3.4 million (approx. RM15.6 million) round from investors including Y Combinator, Gradient, and Lightspeed backs AI agents that do the accounting itself, not just digitize surrounding paperwork. These agents check incoming invoices, assess them for tax, code them, match them to purchase orders, resolve mismatches, and prepare postings into ERP systems like SAP and Microsoft Dynamics.
The practical impact is already visible: Finto reports that the majority of invoices at its first customers now run without a manual human touch. That is not incremental productivity; it is a structural change in how finance teams work. For decades, software digitized everything around accounting while judgment remained human, so finance staff still handled invoices by hand. By putting AI agents directly inside those judgment-heavy workflows, Finto is a clear proof that AI agents enterprise deployments can move beyond analytics dashboards into line-of-business execution. This is the future of AI automation accounting: narrow, deeply integrated agents that do the work rather than suggesting it.
Aidan and Medallia: Customer Service AI Turns Pro
If accounting is the back office test bed, customer-facing workflows are where AI agents will either win trust or fail loudly. Sable’s USD 45 million (approx. RM207 million) raise to expand Aidan shows how far customer service AI is moving from scripted chatbots to interactive, task-completing agents. Aidan conducts customer interactions by handling software, speaking, and even demonstrating products in real time; it can recognize what is on the screen, select interface elements, move through workflows, and explain each step as it goes. Sable is initially targeting qualification, live product demos, onboarding, and account expansion, and the platform is built to run at any time in multiple languages, reducing the need for a human in every conversation.
On the experience platform side, Medallia has already shipped Frontline-Ready AI to move enterprise AI “from proof-of-concept into production,” including Smart Response Drafts that write AI-assisted replies to customer feedback. The company says some customers have cut response times by as much as 80% as a result. After a period of financial uncertainty and a looming equity wipeout, Medallia reset its capital structure and now has USD 150 million (approx. RM690 million) in new capital behind an existing USD 500 million (approx. RM2.3 billion) product commitment, with a third stage focused on conversational experiences and agentic integrations. This is customer service AI evolving into orchestration: agents not only draft responses but trigger workflows across sales, service, and marketing clouds, responding to feedback and taking action without waiting on humans.
AIsa and Medallia: Building the Infrastructure for AI Agent Economies
The most underestimated piece of AI agents enterprise adoption is not clever prompts, but plumbing. AIsa is tackling that head-on with a unified transaction network for AI agents, backed by USD 6.5 million (approx. RM29.9 million) in total funding including a seed round co-led by Alibaba and Tribe Capital. As AI agents move from answering questions to taking actions on behalf of users, they need a way to discover, access, and pay for models, APIs, data, and digital services—yet most of these resources still assume human accounts, contracts, and manual billing.
AIsa’s platform combines a resource gateway with an agentic payment layer that meters usage, enforces spending controls, and settles transactions in fiat or stablecoins through a single programmable interface. Since 2025, it has onboarded more than 50,000 registered agents without paid marketing, with agent users up 150x and transactions up 200x in just four months, and it already ranks as a top seller and server in the x402 ecosystem. With the new enterprise AI funding, AIsa plans to deepen budgets, approvals, and audit trails, and scale infrastructure so autonomous agents can transact at internet scale. In parallel, Medallia’s stage-three roadmap—agentic integrations that move data between its platform and sales or service clouds—shows incumbents converging on the same need: reliable, controllable pipelines for agent-driven workflows.
Why This Funding Wave Matters—and What Comes Next
These four enterprise AI funding moves are not random bets; they mark a clear shift from general-purpose AI tools to specialized AI agents designed for specific business processes. Finto’s accounting agents, Sable’s customer-facing AI employee, AIsa’s transaction fabric, and Medallia’s agentic AI roadmap all share a common thesis: narrow, deeply integrated agents will deliver more value than yet another open-ended chatbot.
The why-now is straightforward. Enterprises have spent years digitizing workflows around human judgment, from finance to customer experience; now AI is reliable enough to move inside those decisions. Medallia’s recapitalization story—shedding debt after a period of renewal anxiety and doubling down on AI-led innovation—shows that customers are demanding more than slideware. The next phase will not be about who has the biggest model, but who can deploy trustworthy agents that plug into ERPs, CRMs, and payment rails with clear controls and audit trails. The risk for enterprises is treating these agents like toys; the opportunity is to treat them like employees with job descriptions, guardrails, and measurable outcomes. Those who design for the latter will set the pace of the AI agent economy.
