PayPal runs 74,000 weekly tasks in Perplexity Enterprise. Teams use it for model validation, channel performance, market trend research, competitive intelligence, and product analysis. “Perplexity gives us the rationale behind every output, and that’s what lets us move with confidence,” says Graham Woods, a model governance lead at PayPal. Read the customer story: https://lnkd.in/g7vQBwR3
Micky Ze
1w
74,000 weekly tasks across 3,700 users is a textbook example of real enterprise AI adoption. That averages out to 4 queries per user, every single work day. It proves the tool has moved from a "nice-to-have novelty" to a core part of the daily workflow. When employees trust the tool enough to use it multiple times a day to unblock their tasks and verify information, the time savings compound fast. Fascinating look at how corporate workflows are shifting right now.
Pixelense
1w
This hits close to home. The "rationale behind every output" line is exactly what separates AI tools that get adopted from ones that get abandoned after a pilot. Confidence in the output matters as much as the output itself. We see the same thing in visual AI — brands don't just want great product images, they want to understand why it looks right. That trust layer is everything.
Des Raj C.
2w
74,000 weekly tasks sounds like scale, but the sharper signal is that research is becoming an operating layer, not a side activity. Once multiple teams start acting on AI-generated rationale, the real challenge becomes consistency of evidence, reviewability, and decision traceability across the company.
Anu Verma
2w
“Rationale behind every output” is where enterprise AI starts separating from casual AI use. The useful shift is not just faster answers, it is giving organizations something they can inspect, contest, and reuse inside actual workflows. Without that, volume is impressive but hard to trust.
Hello everyone please help me money 💸 fast agent 🙏🙏🙏
That point about AI checking AI is spot on. Once you get that loop running, it really changes the game for how teams handle risk. It makes me wonder what the next hurdle will be once verification gets fast enough to keep up with the generation side. Thanks for sharing that perspective, it is a really interesting way to look at where we are headed.
Max Votek
1w
The use-case list is the part most CIOs will copy. Model governance + market intel + competitive analysis in one tool collapses three vendor lines. Internal politics of merging those teams is harder than tool selection.
Like a pattern shift in enterprise AI adoption, success isn’t driven by one tool replacing workflows, but by giving teams verifiable, citation-backed systems that slot into existing decision loops and reduce the cost of trust at scale.
74,000 weekly tasks is a serious signal. The real unlock isn’t just faster research it’s giving teams enough rationale and traceability to actually trust the output at enterprise scale.
Model validation as a use case is the tell. PayPal is using AI to check AI - once that loop closes, the bottleneck shifts from trust to verification throughput