Jason Horman

Clay doubling down on increasing action costs while a lot of gtm-e world is laser focused on BYOK workflows with Claude Code is an interesting tension. In the last ~week Cargo shipped a CLI, Apollo.io released an MCP connector for Claude & most of my feed is Claude Code. The ecosystem energy feels entirely focused on terminal-first. I’ve loved Clay for orchestration but have increasingly felt that the UI/UX is slowing me down as I ramp up my terminal-first workflows. That said, terminal-first isn’t multiplayer yet and isn’t in it's final form - esp for larger companies where multiple humans benefit from a shared UI. What’s interesting is this is coming from the people who coined “GTM Engineer” in the first place. The early adopters Clay built the category for are the ones feeling the pull away from the orchestration layer Clay monetizes. The dispersion is probably temporary. But if Clay’s pricing move take them further from the early adopters who built the category, that’s a deliberate bet to cross the chasm to enterprise, while some subset of builders build elsewhere. Interesting to watch.

17

3 Comments

Your technical debt is worth more than you think. Not because you should keep it. Because of what it tells you before you remove it. Every PE firm, every board, every acquirer walks into diligence and treats tech debt like a liability on the balance sheet. Something to discount. Something to apologize for. They're reading it wrong. Tech debt isn't just rot. It's a fossil record. Every shortcut, every deferred migration, every "we'll fix it later" is a decision someone made under pressure. And those decisions are a map of what the organization actually valued — not what the investor deck said it valued. That monolith that should have been decomposed three years ago? Leadership didn't have the appetite for disruption. The patchwork of acquisitions duct-taped together? M&A ambitions that outpaced integration capacity. The hand-rolled authentication system nobody wants to touch? Someone was afraid to break what was working. Read the debt. Read the org. In my organization, we hunt tech debt relentlessly. My engineers are tasked with removing it. But my job is different — I need to figure out what systemic issue created that debt in the first place and make sure that gets removed too. Kill the debt and kill the factory that produced it. Otherwise you're just mopping the floor with the faucet running. Here's the part nobody says out loud: the companies with zero tech debt should worry you more. Either they're not shipping fast enough, they've been so obsessed with engineering purity they lost the market, or they're lying. Tech debt in the right places is evidence of aggression. It means someone made a bet. The question isn't whether a company has tech debt. It's whether they understand the bets they made and know which ones to unwind. Stop apologizing for your tech debt. Start reading it. Then remove it — and whatever created it. --- Repost if this changes how you think about technical due diligence. If not, well... repost anyway!

15

6 Comments

Hot take: if you're not designing for cache failure modes during development, you're just scheduling a production incident. We're deep in building right now, architecting a backend for burst-heavy read traffic. The kind where thousands of requests hit the same endpoints at once. And here's the thing nobody talks about: the default behavior of most caching setups is a ticking time bomb. One hot key expires → every request bypasses cache → database gets crushed → cascading failures. This is the Thundering Herd problem. Most teams discover it the hard way. We're solving it before we ship: → Request coalescing via Redlock — one request rebuilds, the rest wait → Stale-while-revalidate — milliseconds of slightly old data beats seconds of downtime → Jittered TTLs — no more synchronized expiry stampedes If you're building for scale and want to understand the full picture — stampede prevention, cache-aside patterns, and the tradeoffs involved — this is worth a read: https://lnkd.in/gtvE7xGf Your cache is a contract with your database: "I will protect you." Make sure the contract holds when things get real. #DistributedSystems #SystemDesign #BackendEngineering #Redis #StartupEngineering

40

1 Comment

Most engineers obsess over writing better code. Very few obsess over building a better system around how code gets written. Last year, we started experimenting with a different approach inside our team. Instead of treating the AI as a code assistant, we treated it as an engineering environment we actively shape. And the shift has been dramatic. Here’s the thing. The real unlock wasn’t smarter prompts or a new model. It was building a workflow where: • big tasks don’t start with code, they start with thinking • complex work gets broken into focused agents instead of one overloaded session • every correction becomes a permanent improvement in how the system behaves • nothing ships until someone proves it works, not just says it does • elegance isn’t an afterthought, it’s a habit • bug fixing becomes autonomous, not managerial What this really means is you aren’t just coding faster. You’re engineering in a way where errors stop compounding and improvements start compounding. We’ve been running this playbook at BoloSign and the difference shows up everywhere. Cleaner diffs. Fewer regressions. Faster shipping cycles. A team that trusts the process instead of fighting the tools. If you’re building with AI every day, don’t just upgrade the model. Upgrade the operating system around it. That’s where the real 10x lives.

9

1 Comment