Todd Beaupré

Scripps accelerating agentic AI at scale. Hot-swapable LLMs connected to sophisticated embedded prompts, secure proprietary data, tools, MCP connections and many-many time saving agents now available and many more in the bullpen. Incredible team effort and massive shared learning!

19

7 Comments

To FTP or to sFTP? Been using these technologies for 30+ years, and somehow I find myself still wondering when to use one or the other. The folks at #kiteworks have a nice article: https://gag.gl/oi1wa2 SFTP (Secure File Transfer Protocol) offers a highly secure method for transferring files by using encryption, authentication, and data integrity checks, making it a significant improvement over the older, unencrypted FTP protocol. However, it comes with drawbacks such as increased complexity, slower performance, and limited native support for compliance features like audit logging or encryption at rest. Tools like Kiteworks enhance SFTP by adding security, governance, and compliance capabilities, making it more suitable for regulated environments.

4

In the consumer world, we want a single interface that can write a poem, plan a vacation, and debug code. In that open world, scale is the only strategy that works. But enterprise workflows don’t live in an open world. Most B2B problems like parsing invoices, routing tickets, classifying clauses operate in closed systems. They have well-defined inputs, explicit outputs, and hard failure modes. In my latest article, I argue that the future of enterprise AI isn't always about getting bigger, it's often about model specialization. The most effective architectures I’m seeing are cascading. They use SLMs for routine volume, and escalate to LLMs only when deep reasoning is required. Read the full breakdown below. https://lnkd.in/gkyJbFWb

57

6 Comments

The difference between AI that works and AI that wins is the right partner. As generative AI moves from experimentation into real world production, I have seen firsthand that scale and infrastructure are not enough. At Shutterstock, we have evolved into an end to end AI model training and evaluation partner. We support teams from data sourcing and custom dataset creation to human in the loop feedback and structured performance insights. In this blog, I share more about how we’re helping builders and enterprise teams move from model performance to real-world outcomes. https://lnkd.in/gwrtD33v

9

ICYMI -- Over the weekend, Google announced the Universal Commerce Protocol (UCP) at NRF -- this is an open-source standard that powers Agentic Commerce. Think of it as a 'common language' that allows AI agents to talk directly to a business's backend. This was built to enable commerce and solve for custom catalog management per platform, and fragmented commerce journeys that lead to abandoned carts and frustrated shoppers. Take a look here: https://ucp.dev/

14

Why do some orgs fail to see the ROI on data initiatives? Because they don’t have a *product* mindset. They see data initiatives as one-shot insights, where you deliver an output, and then it’s done. You may be asked to re-run it in future, but that’s tomorrow’s problem. And then all of a sudden it’s being run weekly and you’re scrambling to maintain it! If you think about the data initiatives as building products, you have to consider its lifecycle, how its going be used and by whom, and how it delivers value. You’ll develop it incrementally using DevOps practices so you can be confident in the code and the outputs, and someone with the right skill set will own it in production and keep an eye on it incase things go wrong. Yes, it *may* take a little longer to deliver that first output, and you are committing more resource to “doing it properly”, but you’ll have a data product that you can trust, reuse, and extend; that offers measurable value; and which has far lower maintenance cost. You should start to see that ROI stack up pretty quickly…

16

1 Comment

What if RETAIL AI didn’t just recommend - but actually executes the next best action across your value chain? On January 8, 2026, Microsoft announced new agentic AI capabilities designed to connect merchandising, marketing, store operations and fulfillment into a coordinated “intelligence layer” so teams can anticipate, decide and act faster. Two parts I’m watching closely: a) Copilot Checkout: shoppers can complete purchases discovered in Copilot without being redirected - now available in the U.S., with partners including PayPal, Shopify and Stripe. b) Retail agents: from brand-aligned shopping agents to catalog enrichment and a store operations agent that supports associates with inventory, policies and next-best actions. And for the shop floor: pairing these experiences with Surface Copilot+ PCs brings secure, mobile, on-device AI performance to every shift. https://lnkd.in/dfP8QyPZ

3

Curious about becoming an Agentic Enterprise? Check out our new step-by-step guide! This playbook offers a clear roadmap to integrate AI agents and transform your business. Activate your systems, empower teams, and move faster with confidence. 

5