What We Heard at the Tokyo Enterprise Automation Roundtable
We attended a closed-door roundtable with IT and operations leads from 8 Japanese enterprises. The data sovereignty conversation came up in every single session.
Read article
dodoAI Blog
Thinking Out Loud
On enterprise AI, data sovereignty, and process automation. From the dodoAI team in Osaka.
All Articles
Inside the dodoAI Agent Runtime: Architecture Decisions for On-Prem Deployment
We made some unusual architecture choices to make our agent runtime deployable in air-gapped environments. This post explains the key decisions and the trade-offs we accepted.
Read article
Rule Mapping 101: How We Translate Enterprise Policy Documents into Agent Logic
Getting an AI agent to enforce your approval policies correctly starts with a rule mapping session. Here is what that process looks like and the common gaps we find in every engagement.
Read article
Osaka's Enterprise AI Landscape: Why Manufacturing and Logistics Lead Adoption
Osaka is Japan's business and manufacturing capital, and its enterprise AI adoption patterns differ meaningfully from Tokyo's startup scene. Data sovereignty concerns are front and center here.
Read article
Decision Logging: How AI Agents Create Audit Trails That Hold Up
An AI agent that makes a wrong approval decision is a problem. An AI agent that makes a wrong decision and cannot tell you why is a catastrophe. Here is how we approach decision logging.
Read article
Connecting AI Agents to Legacy ERP: What the Integration Actually Looks Like
ERP integration is where on-prem AI deployments get stuck. Most agents assume REST APIs. Most enterprise ERPs speak SOAP, JDBC, or proprietary connectors. Here is how we handle it.
Read article
Why Compliance Automation Requires Explainability — Not Just Accuracy
A compliance check that says 'approved' or 'rejected' is not enough for a regulated environment. Your auditors will ask why. Your AI should be able to answer.
Read article
On-Premises AI Deployment: A Technical Checklist for Enterprise IT
Before you deploy any AI agent inside your network, these are the infrastructure and security checkpoints that matter. Our own onboarding experience distilled into a practical list.
Read article
Manual Vendor Onboarding vs. Agent-Automated: A Time and Risk Comparison
A typical mid-size enterprise takes 3-6 weeks to onboard a new vendor. We mapped the manual process step-by-step and identified where agents can compress the timeline without reducing diligence.
Read article
APPI and AI Data Handling: What Japan's Enterprises Need to Know
Japan's Act on the Protection of Personal Information has specific implications for AI systems that process employee and vendor data. On-premises deployment is the simplest path to alignment.
Read article
What 'Agentic AI' Actually Means for Enterprise Process Automation
The term 'AI agents' is being applied to everything from chatbots to RPA wrappers. Here is what genuine agentic execution means for back-office processes.
Read article
The Hidden Data Risk in Procurement Automation
When your AI approval system sends purchase orders to a shared cloud model, you are leaking vendor pricing, contract terms, and sourcing strategy. Most companies don't realize this is happening.
Read article
Why Enterprise AI Belongs On-Premises: The Case for Sovereign Deployment
Cloud AI vendors offer convenience. Enterprise IT teams need sovereignty. We explain the trade-off and why we think the industry is heading the wrong direction.
Read article