Your AI Agents Are Fast. Your Process Isn’t. Here’s How to Fix It.

AI agents often stall in workflows where human steps remain unchanged, trapping value at each handoff. Salesforce reports 95% of companies fail to realize AI benefits because isolated automation leaves bottlenecks intact. Even advanced reasoning across decades of data cannot outpace manual processes, limiting return on investment. The company argues agentic transformation requires redesigning entire workflows, not just inserting tools.

To address this, Salesforce developed Rapid Innovation Squads (RIS), small teams combining business stakeholders and agentic builders. Early pilots showed outcomes 30 times faster than traditional methods, with decisions made 10 times quicker. Projects like the Sales Signals Slack app were built in days instead of months, generating 28,000 opportunities. The model prioritizes speed, security, and governance over lengthy project plans.

RIS projects have delivered measurable efficiency gains, including an 80% adoption rate for a Slack Skill that logs 23,000 weekly activities automatically. Legal teams reduced case routing time from 20 minutes to near-instant, saving $350,000 annually with a 70x return. Agentic routing improved accuracy by nearly three times, eliminating duplicate work. These results demonstrate the model’s scalability across departments.

Starting in December, Salesforce will expand RISFest companywide, inviting teams to submit business problems for rapid squad-based solutions. The initiative aims to prove the model’s viability at enterprise scale, emphasizing secure, governed environments. The goal is to shift from task automation to full workflow redesign, ensuring every step operates at agentic speed.