Notes from
the work.
Practical writing on AI agents, automation, integrations, and the operating decisions around them — what to build, what to skip, and how to tell the difference.
Decisions, systems,
and the rules around them.
12 articles. Each one is written to answer a question a buyer actually asks, and reviewed before it is published.
AI Agent vs. Chatbot vs. Deterministic Automation: A Decision Framework
Choose the smallest system that fits the work by comparing ambiguity, authority, evidence, failure cost, and operating ownership.
· 12 min read
AI Automation Cost and ROI: A Baseline-to-Decision Worksheet
Calculate total ownership cost, realistic benefit, uncertainty, and failure conditions before approving an AI automation project.
· 13 min read
Small-Business Cybersecurity: A Risk-Based Starting Plan
How to prioritize identities, updates, backups, logging, vendors, and incident readiness without claiming perfect protection.
· 10 min read
How to Handle Business Data in Generative AI Tools
A control framework for approved tools, data classification, retention, prompt injection, access, and incident response.
· 10 min read
When a Small Language Model Fits a Business Workflow
How to compare a smaller model with hosted and larger alternatives using task evidence, deployment constraints, and total ownership.
· 9 min read
When Fine-Tuning an AI Model Is Worth Considering
A decision guide for comparing fine-tuning with prompting, retrieval, tools, and conventional software rules.
· 9 min read
How to Choose a Business Process for AI Automation
A practical way to select, control, and measure an automation project before expanding its scope.
· 9 min read
Evaluating Open-Source AI for Business Use
A practical review of licenses, deployment control, security, operations, and total ownership before adopting an open model or framework.
· 9 min read
RLHF, DPO, and Preference Tuning: A Business Guide
What preference-based post-training can change, what it cannot prove, and when a business should avoid owning the training loop.
· 10 min read
How to Choose an AI Model for a Business Workflow
A vendor-neutral evaluation method based on task quality, risk, latency, integration, and total operating cost.
· 10 min read AI Agent Workflows: Scope, Controls, and Release Criteria
How to decide whether a workflow needs an agent and how to limit tools, permissions, and failure modes.
· 10 min read How to Evaluate Whether an AI Investment Is Worth Pursuing
An executive framework for separating hypotheses, measured evidence, operating costs, and decision risk.
· 9 min readNothing under that topic yet. Ask us the question directly