Where to Place Your First AI Agent in Accounting
Your first agent shouldn't live inside your accounting platform. Start where the door is already open.
Blog
Insights, guides, and updates
The McKinsey AI leadership gap
Accenture just cut 11,000 and trained 70,000 in AI — what CEO Julie Sweet called "exiting people where reskilling isn't viable." McKinsey's data says only 14% of firms have leaders like that. If your practice's AI investment isn't producing results, the bottleneck isn't your tools or your team. It's you.
Are you using AI to solve a cost problem or a revenue problem?
Most CAS practices are using AI to solve a cost problem. PwC's April research shows that's the 80% position — and it caps your upside. Here's what the 20% are doing instead, and why your niche is the prerequisite for getting there.
Weekly AI Roundup: The deferral window just closed
PwC and Accenture both restructured around AI this week. Anthropic shipped Opus 4.7 and quietly moved enterprise billing to per-token pricing. OpenAI bought its way into personal finance. Here's what a week like this means if you're running a CAS practice.
You're five minutes away from your first AI agent
2026 is the year of the agent — and you're five minutes away from building one you could reuse every day. An agent isn't a new technology. It's the same AI you're already using with a better job description. The gap between a prompt, a workflow, and an agent is smaller than most CAS practitioners think.
What does your practice sell when the platform does the work?
A major accounting platform just guaranteed 95% automation or you don't pay. When the software handles nearly all the transactional work, the question isn't whether your bookkeeping fees survive — it's what your practice actually sells next. The answer reshapes your cost structure, your fee model, and your client service.