More harm than good

In tech circles — especially here in Silicon Valley — there’s a pervasive narrative around AI that I believe does more harm than good. Right now, SaaStr is leading the charge.

“AI is coming for your job. Prepare for automation. Fewer people. More code. Better margins.” - Jason Lemkin on LinkedIn

We reject that idea. We reject it not because AI can’t automate certain tasks. Of course it can. But because using AI solely to replace people is the most short-sighted way to deploy it — and the least effective way to build a company that thrives.

🛠️ The Real Value of AI: Work Reimagined

We don’t see AI as a replacement engine. We see it as a transformation engine.

That belief is core to how we’re building our platform — from DeliveryGPT, our generative delivery intelligence engine, to our prompt-based tools for fulfillment strategy and conversion rate optimization.

We’re giving humans in the loop better levers to make better decisions faster, both at Fenix and for our customers

🧠 Real-World Example: Empowering Teams at a Major Apparel Brand

One of our leading customers — a household name in the premium apparel space — is using our beta access to prompt-based evaluations of their carrier network across states like California, Texas, and Florida.

Here’s what changed:

Before: Their operations analysts had to dig through dashboards, export CSVs, and wait on BI to understand which regions had the most delayed deliveries and which upgrades drove conversion. It was reactive. It was manual. It was slow. 

Now: With DeliveryGPT, those same analysts can ask

“What happens to revenue per visitor if we improve delivery time by 1 day in Southern California?”

They get results instantly — but they’re not just passive recipients. They interrogate the data, pressure-test assumptions, and even refine the prompts to make sure the answers align with real-world fulfillment constraints.

That’s human-in-the-loop in action: AI as a thinking partner, not a shortcut.

Get Insights for Your Brand

Talk to Our Consultant

Get notified about our upcoming blogs

Follow us

📊 A Shift in Metrics: From Dashboards to Data Design 

Here’s a concrete example.

Traditionally, a BI analyst or front-end engineer at an eCommerce brand might be evaluated on: 

• Number of dashboards delivered
• Quality of visualizations
• Load times or click-through rates

But we’re entering a new world — one where users query fulfillment data in natural language and get instant answers. Dashboards are being replaced by prompt-native experiences.

So what happens to the BI team?

We believe their value goes up, not down. Because now, their success looks like:

• Building prompt-ready data models
• Structuring manifests so the AI knows what to return
• Designing visual guardrails to prevent misinterpretation
• Enabling decision-makers to explore insights confidently, not just passively consume charts

That’s not about elimination. That’s about elevation.

🔁 What We’re Doing at Fenix

Internally and with our customers, we’re doing a few things differently:

• We measure performance not by what got automated — but by how we empowered teams to make faster, better, more contextual decisions
• We build tools that make complex operations conversational — not tools that pretend complexity doesn’t exist
• We view AI as a partner, not a product. Something that learns with us, adapts with us, and amplifies the best of what our people already bring

🚀 Final Thought: This Isn’t Just Philosophy — It’s Product Strategy

The future of delivery, fulfillment, and conversion optimization won’t be shaped by the companies who eliminate the most roles

It’ll be shaped by those who give their teams superpowers — through GenAI, through data modeling, and through bold rethinking of what “good work” looks like.

At Fenix, that’s the future we’re building.

Not AI instead of humans. AI alongside humans