Hosted by TrueWatch with AWS as supporting partner, AWS Observability Day Taipei 2026 took place on August 11 at the AWS Taiwan office, drawing CIOs, CTOs, and technology leaders from across the industry to explore how AI Agents are bringing tangible change to cloud operations and observability. The event centered on "How AI Agents can be applied to cloud operations and observability," featuring both AWS's own automation practices and the newly unveiled Toby AI product line from TrueWatch — offering a comprehensive look at how far the industry has come over the past year.
AWS Shares Automation in Practice: From Amazon CloudWatch to AWS DevOps Agent
In a session titled "Move Beyond Reactive: Transform Cloud Ops with AWS DevOps Agent," AWS shared how enterprises can build on the cloud monitoring and alerting foundation established by Amazon CloudWatch and further adopt AWS DevOps Agent — helping operations teams shift from a reactive model of "wait for the alert, then respond" toward a proactive approach that detects anomalous trends early and automates remediation, accelerating the response time from detection to resolution.
TrueWatch Unveils the Toby AI Dual Engine: TruePilot and Agents
TrueWatch's technical team officially introduced its Toby AI product line, comprising two complementary modes:
-
Toby AI TruePilot (Assist mode): A context-aware AI assistant built directly into the TrueWatch interface. It generates charts, writes query syntax, and investigates root causes using natural language — no query-language expertise required — and never touches production systems, lowering the barrier to adopting the platform.
-
Toby AI Agents (Execute mode): Fully autonomous AI operators that can be configured into dedicated agents for roles such as SRE, Application, FinOps, and Security. They monitor, analyze, and execute tasks around the clock, and support agent-to-agent (A2A) collaboration as well as interoperability with third-party agents.
TrueWatch's technical team demonstrated a real-world incident scenario on stage: the same alert that once took an average of 47 minutes to diagnose and resolve now takes just 5 minutes with Toby AI — a nearly 9.4x improvement in efficiency. The team was careful to emphasize that AI Agents are not meant to replace engineers' judgment: any potentially destructive action, such as a rollback or deletion, still requires human approval, keeping final authority firmly in the hands of the engineering team.
Workshop in Action: Letting AI Find the Root Cause in a Real, Broken Environment
In the afternoon session, attendees stepped into a hands-on workshop built around a genuinely broken AWS EKS environment (loaded with a dataset of one million articles) — with no hints given about where the problem actually was. SREs, operations engineers, and technology leaders in attendance worked through the environment using Toby AI TruePilot and Toby AI Agents to conduct root cause analysis stage by stage, tracking down performance bottlenecks and the underlying issues. Toby AI pinpointed an unindexed database query as the culprit and executed the fix on the spot, cutting the homepage load time from roughly 10 seconds to about 1 second — giving attendees a firsthand sense that "AI isn't there to replace troubleshooting, it's there to accelerate it.
A TrueWatch Perspective: The Gap Between AI's Promise and Its Practice Is More Real Than It Looks
Drawing on its ongoing observation of how Taiwanese enterprises are adopting AI for operations, TrueWatch notes that "too many disconnected tools, too hard to operationalize AI" isn't an isolated complaint from a handful of companies — it's a challenge the broader market is grappling with together. Within Agentic AI use cases, ITSM/AIOps already ranks among the leading areas of enterprise adoption. Yet when it comes to actual implementation, "lack of a cross-system, integrated view" and "difficulty evaluating and operationalizing AI/automation" remain the two obstacles technology leaders raise most consistently — followed by execution-level pain points such as alert overload making prioritization difficult, incidents taking too long to diagnose, and insufficient headcount to respond in real time. TrueWatch believes this is precisely the problem its Toby AI dual-engine architecture is designed to solve: not by adding yet another tool, but by consolidating scattered signals, judgment, and action into a single, trustworthy interface.
About TrueWatch
TrueWatch is an observability company serving the international market, helping enterprises monitor complex environments, catch problems early, and keep operations stable and reliable. TrueWatch employs more than 200 people globally, serves over 500 customers worldwide, and its platform supports 650+ integrations that can be collected through a single agent.
TrueWatch is an AWS ISV Partner and was named a 2025 AWS Marketplace Partner of the Year (GCR). It is also officially listed on AWS Marketplace, allowing AWS customers to adopt the TrueWatch platform directly through their existing procurement workflows. That depth of partnership extends to local presence as well — TrueWatch maintains a dedicated engineering and SRE team based in Taiwan, making it one of the few observability platforms with both local business and engineering teams on the ground in Taiwan.
Media Contact
TrueWatch Marketing Team [email protected] Amazon Web Services, AWS, and related marks are trademarks of Amazon.com, Inc. or its affiliates.