Sonika Janagill·Jul 8The best way to learn AI? Teach it.How leading the WCC AI Learning Series taught me more than it taught anyone else, and why you should try it too.A response icon1A response icon1
InGoogle Cloud - CommunitybySonika Janagill·Apr 25Instructions. Skills. Tools. How Google Embedded Skills Into Every Layer of Its Agent StackPrompt Bloat has a name and a fix. Skills are now load-bearing across Google’s agent stack: from on-device Gemma 4 to enterprise Gemini…
InGoogle Cloud - CommunitybySonika Janagill·Apr 22Google Cloud’s Agent Ops Stack: Why Deployment Is No Longer the Hard PartThe Gemini Enterprise Agent Platform shifts the conversation from deployment to governance, from the Request/Response era to the Long-Lived…
InGoogle Cloud - CommunitybySonika Janagill·Apr 5Gemma 4 and the Rise of Agentic CommerceNative multimodality, configurable thinking, reliable function calling, and true Apache 2.0 licensing: why Gemma 4 is the most…
Sonika Janagill·Mar 26The Skills Explosion Is Here. Enterprise Governance Isn’t.Enterprise Difficulties in Managing & Securely Using Agent Skills
Sonika Janagill·Mar 16AdkBot: Building a Personal AI Agent with Google ADK on Cloud RunCloud-native, 6-layer GCP security, human-in-the-loop approvals, a truly secure personal AI agent in 2026
Sonika Janagill·Dec 5, 2025Building the Internet of Agents: AGNTCY, Protocols, and Mach Alliance’s Agent EcosystemEnterprises are racing to build with AI agents. But the real risk isn’t picking the wrong model — it’s building on platforms that turn…
Sonika Janagill·Nov 30, 2025Implementing Zero-Trust Multi-Cloud: A Complete WIF Setup GuideStep-by-step guide to implement Workload Identity Federation for AWS→Vertex AI and Azure→Vertex AI.
Sonika Janagill·Nov 26, 2025Stop Storing Cloud Secrets: How Workload Identity Federation Eliminates Static KeysThe most dangerous part of modern ML isn’t model accuracy. It’s the secrets we scatter across systems like digital landmines.
Sonika Janagill·Nov 8, 2025The Hidden Cost of Data Chaos in ML ProjectsReal numbers behind ML project failures and the business case for MLOps architecture that actually worksA response icon1A response icon1
