Two recent developments point in the same direction for firms that want practical AI: services software is moving toward systems that understand context and take action, while legal tech is focusing on repeatable back-office workflows that can be governed and audited. For law and accounting leaders, the message is simple: start with the work that repeats, then build AI around it.
The market is moving beyond chatbots
Certinia's new System of Action is designed to connect professional services, customer success, financial management, and CRM data, then apply judgment and act on trusted data. That matters because the reported gap is not AI adoption itself, but AI that delivers on expectations.
For firm leaders, that is a useful signal. Generic assistants can answer questions, but custom AI earns its place when it helps a firm move from information to action across sales, delivery, finance, and client service.
Back-office automation is becoming the first practical win
Caddi's approach for lawyers is centered on discovering repetitive back-office work, then automating it across the tools a firm already uses. Its workflow starts with observing what repeats, building the process over screen share, and then governing who can build, approve, and run it.
That model is especially relevant for law and accounting firms because many of the easiest AI wins live behind the scenes: billing support, matter or engagement setup, document handling, CRM updates, and routine follow-up. Those tasks are often easier to standardize than client-facing advisory work.
What this means for custom AI in a professional-services firm
The common thread in the news is not flashy AI; it is structured automation with controls. Firms that want results should think in terms of workflows, approval paths, auditability, and ownership rather than one-off prompts.
That is also where agentic workflows can help. If an AI system can identify a repeatable process, execute steps across systems, and log what happened, it becomes easier to create measurable operational leverage without giving up oversight.
A practical starting point for law and accounting leaders
The best first use case is usually the process people already do the same way every week and complain about every month. That may be client intake, billing cleanup, engagement setup, document triage, or internal routing.
Build from there with clear governance. Decide who approves the workflow, what data it can touch, how exceptions are handled, and how success will be measured. That approach is more durable than chasing broad AI adoption for its own sake.
- Look for AI that connects data and acts on a defined workflow, not just software that answers questions.
- Start with repetitive back-office processes where the firm can standardize steps and measure results.
- Put governance in place early so approvals, ownership, and audit trails are part of the workflow.
- Use custom AI to reduce manual coordination across systems, not to replace professional judgment.
Sources watched
- Certinia Rolls Out System of Action for Services Businesses, Expands Veda AI (CPA Practice Advisor AI)
- Q&A: Jen Leonard on Unprecedented and the Future of Law (Artificial Lawyer)
- Caddi - Back-Office Automation For Lawyers (Artificial Lawyer)
