The successful implementation of AI agents requires more than the agent itself.
This should come as no surprise to anyone who really understands the implementation of tools in the workplace: *any* tool implementation involves much more than the tool itself. Far too many companies, however,
The article [One Year of Agentic AI: Six Lessons from the People Doing the Work](https://www.mckinsey.com/capabilities/quantumblack/our-insights/one-year-of-agentic-ai-six-lessons-from-the-people-doing-the-work?stcr=01EDC1D68F604923BC43281A965B0491) suggests that many projects are failing - companies are dismantling attempted agent deployments. It claims success requires more than the agent itself — it demands redesign of processes, trust, governance, and feedback mechanisms. It sets out a practical analysis of implementing AI:
1. **It’s not about the agent; it’s about the workflow**
– Focus on redesigning end-to-end processes, not on building isolated agents.
– Map workflows, identify pain points, and integrate agents where they reduce friction.
– Example: legal provider logged user edits in contract review to refine the agent’s knowledge base.
2. **Agents aren’t always the answer**
– Some tasks (low variance, highly standardised) are better handled by simpler automation or prompting.
– Evaluate variance, decision complexity, and error profiles before building agents.
3. **Stop “AI slop”: invest in evaluation and trust**
– Users lose confidence quickly if outputs are poor or inconsistent.
– Treat agents like new employees: define roles, onboard them, evaluate continuously.
– Use frameworks to monitor precision, recall, hallucination rates, and alignment with human judgment.
4. **Track and verify every step**
– End-outcome metrics are insufficient; observability into intermediate steps is essential.
– Logging, error detection, and feedback loops allow early diagnosis and refinement.
– Example: in document review, monitoring revealed upstream data quality issues driving errors.
5. **The best use case is the reuse case**
– Avoid duplicating effort by creating modular, reusable components (prompt templates, evaluation logic, retrieval tools).
– Shared platforms can cut nonessential work by 30–50%.
6. **Humans remain essential**
– Agents can execute many steps, but humans must validate, handle exceptions, and oversee performance.
– Clear workflow design should define handoffs between humans and agents.
– Strong UI features (highlighting, click-to-validate) build user trust and adoption.