Insights
Notes From the Work
Practical writing on the problems we run into building and operating contact centers — the sequencing decisions, the metrics that mislead, and the engineering that separates a pilot from production.
- AI Agents5 min read
AI Agents vs. Chatbots: The Operational Differences That Decide Adoption
The difference between an AI agent and a chatbot is operational, not technical. Here are the differences that decide whether customers actually use them.
- Healthcare4 min read
The Contact Center Is Healthcare's Real Front Door
Patients meet your health system through a phone call long before a lobby. Designing healthcare's real front door: governance, authentication, safe automation.
- Credit Unions4 min read
Credit Unions Win on Service — Until the Phone Rings
Credit unions promise personal service, but the contact center often hides the relationship. How small teams modernize to keep the membership promise at scale.
- Strategy4 min read
How to Rescue a Transformation Program That Has Lost the Room
A practitioner’s playbook for transformation program rescue: stop the clock, rebuild belief with one certain deliverable, re-baseline in public, and fix the system that caused the stall.
- Strategy4 min read
The Questions to Ask Before You Sign Any CCaaS Contract
An operator’s guide to CCaaS vendor evaluation: exit terms, usage-based pricing traps, implementation ownership, and the questions demos are designed to avoid.
- AI Agents8 min read
Contact Center AI Readiness: What Has to Be True Before AI Will Work
Most contact center AI programs do not fail because of the model. They fail because the environment around the model was never ready. Here is the readiness test we apply before recommending AI to anyone.
- Amazon Connect7 min read
Sequencing an Amazon Connect Migration So No Single Step Can Hurt You
Most contact center migrations fail at the cutover, not the build. The fix is sequencing — choosing an order of migration where every phase is small enough to reverse.
- AI Agents6 min read
Containment Is the Wrong Metric for Self-Service
Containment counts customers who gave up alongside customers you helped. If it is your headline self-service metric, you are optimizing for the wrong outcome.
- Agentic AI8 min read
What Agentic AI Needs Before It Goes Anywhere Near Production
Building an agent that acts is straightforward. Building one you can operate, audit, and stop is the engineering problem — and it is the part most pilots skip.
Topics we write about
- Amazon Connect
- CCaaS
- CX Strategy
- AI Agents
- Agentic AI
- Generative AI
- Workforce Management
- Customer Engagement
- Architecture
- Product Engineering
Want this applied to your operation?
The writing is general. The engagement is not — we start from your contact mix and constraints.