AI agent infrastructure becomes the new backbone for autonomous software deployment
AI agent infrastructure refers to the monitoring, governance, and cloud systems that let enterprises deploy autonomous software at scale while keeping it observable, secure, and accountable across live business environments. The latest funding rounds show this stack moving from experiment to core IT priority. Coralogix’s USD 200 million (approx. RM920 million) Series F for agentic AI monitoring signals that observability is now essential, not optional, as enterprises replace dashboards with AI-driven incident investigation. Governance is catching up fast: Willow’s seed round and Opal Security’s USD 23 million (approx. RM106 million) raise target the growing risks from agents that can log in, move data, and trigger workflows like human users. Alongside them, cloud platforms such as Niteshift and email-native products like Upstream show how enterprise AI agents are being wired directly into day-to-day tools, from codebases to inboxes.
Coralogix and the rise of agentic AI monitoring in production
Coralogix’s USD 200 million (approx. RM920 million) Series F round, led by Advent and the Canada Pension Plan Investment Board, is a direct bet on agentic AI monitoring as enterprises scale autonomous software deployment. The company reports more than 60% revenue growth and around 30 customers spending over USD 1 million (approx. RM4.6 million) annually, underscoring demand for observability tailored to AI agents. More than half of its enterprise customers already use Coralogix’s agent Olly or their own models to investigate incidents via command line, “gradually eroding the traditional dashboard” in the words of CEO Ariel Assaraf. Instead of staring at charts, operators ask agents questions about logs, metrics, and traces. This shift turns monitoring into a conversational interface and makes observability a critical layer of AI agent infrastructure, where failures or misbehavior can be detected and explained before they damage operations.
AI governance platforms move to control access, identity, and risk
As AI agents connect to sensitive systems and act with fewer human checks, AI governance platforms are becoming central to security strategy. Willow, founded by former Wix engineers, raised USD 7 million (approx. RM32 million) in seed funding to give companies granular control over how agents access internal tools such as Claude, ChatGPT, Cursor, Gemini, and Codex. According to Willow, 79% of companies have introduced AI agents and 73% run multi-agent systems, yet 65% reported agent-related incidents in the past 12 months. Opal Security extends this control to identity, managing human, service, and agentic AI credentials in one access graph. Its USD 23 million (approx. RM106 million) round and customers like Databricks, which runs 86,000 just-in-time access requests through the platform, highlight how agent identities must be governed with the same discipline as people to contain blast radius and support safe autonomous software deployment.
Cloud platforms retool infrastructure for coding and email-native enterprise AI agents
Beneath governance and monitoring, a new layer of AI agent infrastructure is reshaping everyday tools. Niteshift raised USD 7 million (approx. RM32 million) in seed funding to provide a full-stack cloud platform for AI coding agents like Claude Code and Codex, running them in fully configured development environments with runtime, services, and verification workflows. By letting teams trigger agents from Slack, Linear, or GitHub and run many sessions in parallel, Niteshift turns coding agents into scalable, production-grade contributors. Upstream, backed by Y Combinator with USD 3 million (approx. RM13.8 million), rebuilds email as a native surface for agents and humans, instead of bolting assistants onto legacy inboxes. Drawing on calendars, notes, and knowledge bases, Upstream’s email-native AI agents draft replies, schedule meetings, find receipts, and act autonomously, with users reporting average savings of two hours per day.
Enterprise AI agent platforms tackle cross-company workflows and specialized operations
On top of these foundations, enterprise AI agents are moving from isolated copilots to systems that coordinate work across teams and even companies. ArchAstro, launched by veterans from Microsoft, Stripe, Statsig, and Meta, focuses on cross-company AI agents that automate complex software deployments and integrations. Its “Forward Deployed Agents” connect distinct corporate environments through code-enforced, privacy-aware runtimes that run acceptance tests across both sides instead of shifting raw data, reducing leakage risk. In parallel, platforms such as fonio.ai, Sandstone, and others are building department-specific and cross-company workflow automations, from customer account management to migrations and bug fixes. Together, these efforts show why investors are backing AI agent infrastructure: monitoring, governance, and runtime platforms are converging into a stack that lets enterprises trust autonomous software with critical operations, not only within their walls but across their partner networks.
