Brantley Davidson

The best AI project I've seen this year isn't about efficiency. It's a firm building a tool to capture 40 years of regulatory expertise before their guy retires. Every mid-market company has this person. The one who just knows. The one everyone calls. AI won't replace them. But it can make sure what they know doesn't walk out the door.

@MatznerJon This is absolutely brilliant

I tap my phone to a sticker on my office file cabinet. Thirty seconds later, my printer spits out a piece of paper with my customized up to the minute morning brief on it. Weather. Calendar. How I'm tracking against my goals. Inbox triage. Update on my Getting Things Done lists. I grab a cup of coffee, walk outside, and read it. No phone. No screen. No notification pile-up. That sticker cost me about a penny. What's actually happening NFC tags. You can buy 50 of them on Amazon for ten bucks. They've been sitting in my file cabinet for two years because I couldn't find a use that justified pulling them out. I finally found one. The tags pair with Apple Shortcuts. Tap the phone, fire a shortcut. Standard stuff. Here's the new part — the shortcut sends an iMessage to my agent. My agent is just Claude Code running in a terminal session on my Mac mini, with access to Obsidian and a bunch of tools I've given it. So the tap becomes: "Run my morning brief." The agent goes and grabs everything I've taught it to grab (via a skill file). Compiles it. Sends it to my printer. I read it on paper. With coffee. Outside. Why not just put it on a schedule? Because I don't want it. I don't want my morning brief showing up at 6 AM whether I'm ready or not. I don't want a notification pulling me back to a screen first thing. I don't want rigid automation dictating when my day starts. The tap is the consent. When I'm ready, I tap. The agent runs. The brief prints. That's it. I have a second tag for my weekly GTD review. A third one to catch me up on inboxes. There's also a tag stuck to the printer itself — I scribble notes on the printed sheet, drop it back on the printer, tap that tag, and the whole thing gets scanned and fed back to my agent. Paper in. Paper out. Digital in the middle. This was todays… with a few redactions :) The bigger thing This isn't about NFC tags. NFC tags are stupid little stickers. It's about how you interface with an agent. Right now, most people interacting with AI are typing into a chat window. That's fine for you, the architect. You know what to ask. You know what NOT to ask. But imagine giving every guy in your warehouse a phone number where they can willy-nilly text an agent with full tool access. Read Write to your CRM. TERRIFYING. They could do some really crazy shit. The agent has the keys to the castle and the interface has zero guardrails. A vending machine doesn't ask you what you want. It has buttons! NFC tags are buttons. You can stick a tag on a piece of equipment. Tap to check it out. Tap to log a problem. Tap to take a screenshot of its current state and file it somewhere. The agent does the actual work — but the interface is bounded to exactly what you want that person doing. That's a fundamentally different shape than "here's a chat window, godspeed." What I'm noodling on I'm going to spend a lot of time on this. Physical doorways into agent flows is where some interesting problems live. Not because NFC is magic. It isn't. It's an old technology. But because most people thinking about agents are stuck in the chat box. The chat box is one interface. The world is full of other ones — buttons, tags, cards, printers, scanners, doors, kiosks — and your agent doesn't care which one it's wired to. Blend the offline and the online. Build the interface that fits the work. Some things I want to do on a screen. Some things I want to do on paper, with coffee, outside. The tag lets me choose.

I didn't set out to run an AI company. Last July I worked an issue with my Vistage group. The question was simple: should I pivot everything Prometheus does to be about AI? We started out helping mid-market companies with their CRM and ERP investments. Then our clients kept asking us how to implement what we were doing for ourselves. At some point it snowballed. I'm sort of an accidental AI company. The biggest pivot of my career was decided in a peer-group room, not a board meeting.

Two different clients this week, completely different industries, asked me the same question with different words. One said: "You're saying top two to three bets. Will there be more? Like a menu?" The other said: "My idea is to tell you shoot for the moon, then you tell us what's realistic." The product mid-market buyers are actually paying for in an AI discovery isn't a roadmap. It's a forced ranking. The line between the two bets we're making and the eight we're killing.

Two different clients this week, completely different industries, asked me the same question with different words. One said: "You're saying top two to three bets. Will there be more? Like a menu?" The other said: "My idea is to tell you shoot for the moon, then you tell us what's realistic." The product mid-market buyers are actually paying for in an AI discovery isn't a roadmap. It's a forced ranking. The line between the two bets we're making and the eight we're killing.

What AI adoption actually looks like inside a mid-market services business in May 2026: The coaching team is using free-tier Claude on personal accounts. The dev team has Claude Pro. The writers have nothing. The CFO is pushing Copilot. The CEO is trying to find the through-line. Nobody called it an AI strategy. It's a knife fight over seats happening one floor below the leadership team's strategy conversation. And the workaround the front line built on their own is more sophisticated than the procurement decision the C-suite is debating.

What AI adoption actually looks like inside a mid-market services business in May 2026: The coaching team is using free-tier Claude on personal accounts. The dev team has Claude Pro. The writers have nothing. The CFO is pushing Copilot. The CEO is trying to find the through-line. Nobody called it an AI strategy. It's a knife fight over seats happening one floor below the leadership team's strategy conversation. And the workaround the front line built on their own is more sophisticated than the procurement decision the C-suite is debating.

@vitaliidodonov simple and dope - 100% agreed

The hiring framework I swear by:

@businessbarista Hype vs reality is huge. Because some people still just (including the leaders) need education on the new foundations (e.g. skills) to begin to understand what hype could look like for them. I haven't found the literacy ladder to exist yet for non-technical leaders with AI.

I spoke to five Fortune 2000 execs today about the state of AI. I asked each one “What’s the most challenging part about this moment in AI?” The CISO said: “There is an ocean-sized gap between hype and reality, which makes discerning what’s real exhausting.” The VP of AI engineering said: “Everyone acts like they’re an expert, yet the main reason so few AI use cases have reached production in enterprises is because true expertise requires experience in scaled systems, enterprise politics, AI fluency, governance and guardrails, and deep process knowledge. Almost no one is actually an expert.” The CTO said: “Our remit is to cut costs, but you can’t actually take AI transformation seriously without increasing AI/R&D budgets up front to ultimately drive bottom line once things are in production and performant. It’s an unrealistic expectation.” The Chief of Staff said: “My job is to drive AI upskilling across the organization, and after doing it for 2 years I’m exhausted. Yes there’s potential ROI from all of the agentic workflows we’re building, but soul and humanity are being sucked out of our processes.” The Finance leader said: “We acquired a multibillion dollar old school business. Getting that business to be AI-native is incredibly painful largely because people aren’t ready or willing to adopt it.” I’m having convos like this every day because I'm building an invite-only AI community for enterprise execs (and interviewing folks before I let them in), but if you find these notes helpful I’m happy to keep sharing them!

Had a prospect call last week where the buyer wanted us to pitch a fully autonomous AI SDR. I told him we don't do that. We've built agents that extend SDR teams. The reps still review, still personalize, still push send. The AI handles the volume they can't get to. His reaction: "Wait, so a human is still involved?" Yeah. That's sort of the whole point. Every AI agent we've built, I still want a human pushing send. That's not a limitation. That's the product.

Had a prospect call last week where the buyer wanted us to pitch a fully autonomous AI SDR. I told him we don't do that. We've built agents that extend SDR teams. The reps still review, still personalize, still push send. The AI handles the volume they can't get to. His reaction: "Wait, so a human is still involved?" Yeah. That's sort of the whole point. Every AI agent we've built, I still want a human pushing send. That's not a limitation. That's the product.

@levie Some of the best AI deployments we've done are ones where the end users don't even know they're using AI. C-suite gets real data about what their customers are saying, what their salespeople are seeing...

Agents are quickly moving from coding to the rest of knowledge work. But to do this we need ways of bridging the advanced capabilities of the AI models with the real-life workflows in the enterprise, by industry and line of business. The models will remain general purpose, but we’ll move to ways of aligning agents to the unique work that gets done in legal, financial services, insurance, healthcare, life sciences, and more. Each industry has its own set of workflows, domain specific context, and data sources that agents need to have access to and be familiar with. Claude just launched an updated set of plugins and skills for the legal industry, including Box. You can now take any of your enterprise contracts or documents and securely work with them in a headless fashion via Claude in a legal workflow. This is just the start of what industry-specific adoption of AI will look like, and equally shows what the future of headless software interaction will look like in the future.

We ran an AI discovery this week where the entire program stalled because the client's partners only had data feeds turned on some of their assets. The most sophisticated AI in the world doesn't matter if the input layer is dark. Mid-market companies hit this wall constantly and need people who understand objectives and tech together to solve these opportunities.

Forward deployed engineers, or equivalent, are about to become one of the most in-demand jobs in tech. And one of the most important functions for AI rollouts. Deploying agents is far more technical of a task than most people realize, often far more involved than deploying software. Software generally works the same way every time, and generally for the past few decades has been updated versions of an existing technology or concept (which basically means easier for the enterprise to update their workflows on a newer system). With agents, you’re actually deploying the equivalent of work output within the enterprise. The customer is effectively using you as a professional services provider for a task, which they expect to get solved nearly end-to-end now. This means you need to actually deeply understand the business process as a vendor, and get the customer from the current to the end state seamlessly. Companies need help figuring out which models will work best for their workflows, they need extensive evals setup often, they need change management support for workflows, they need to get their data setup for the agents, and constant tuning of the agentic system for their process. Massive role in tech now. And another example of the kind of highly technical work that AI is creating.

@IAmPascio AI

If you don't have Stanley for 𝕏™ (AI Head of Content) yet... Like and comment “AI” and I’ll send the link to you.

@Jason I’m in

We started an AI founder twitter group... reply with "I'm in" if you're a founder and want to be added

@_TrueVoodoo Trust me they slap

Coming soon. Very soon.

@shiftj We are doing this for middle market - primary ICP manufacturing, home services, logistics

Massive trend I've seen in YC companies as well. The demand for internal AI transformations seems to be endless. Anyone working on an agency play here to fill the gap?

@rushicrypto Today thought the same that newspapers could be back

I think AI will flood the internet with so much fake and useless content that people will slowly stop trusting it and go back to living real life like it’s the 90s.

@polyaivoice PolyAI

PolyAI has raised $200M from Nvidia, Khosla Ventures, and multiple top VCs. We're one of the fastest-growing companies in the UK, and we handle 500M+ calls for: • Marriott • PG&E • Gordon Ramsay's restaurants • And 3,000 more real deployments Which means that if you've ever called them, chances are you've talked to our voice agents. Every restaurant we onboard books thousands in revenue within 30 days. But how? Because PolyAI works 24/7, answering every call in <2 seconds, and we also: • switch between 45+ languages • handle payments & cancellations • verify identities • and even upsell your services If you want to try creating an agent with PolyAI, we built Agent Studio Lite to make it easy. Just enter any URL, and in 5 minutes it will analyze your website and build a working agent. We're opening early access to a limited number of people. Comment "PolyAI" and we'll add you to the waitlist and give you 3 months for free!

Quick question for business owners: What's the most time-wasting manual process in your business right now? (Trying to see if there's a pattern in what actually needs automation vs. what people THINK needs AI.)