A recap of our workshop with Golf.dev CEO Wojciech Błaszak on building with MCP in production, covering tool design, remote server adoption, customer support use cases, and what actually breaks in the wild.
An intimate dinner in SF for engineers building LLM-native tools. We'll talk shop about Claude, Cursor, and custom workflows, over Thai food, of course.
Join us with Golf.dev CEO Wojciech Blaszak to explore how support teams can use MCP to build secure, resilient AI agents—covering real-world workflows, failure handling, and security best practices.
We were honored to be recognized in Mux’s recent post on AI customer support. Grateful for the shoutout, and excited to keep building thoughtful, configurable AI systems for support teams.
A practical guide to building lightweight QA systems that keep your AI agents sharp, safe, and on-brand.
Engineers, caffeine, and clicky keys. Our latest event proved that typing fast is still a core skill, especially when there’s a leaderboard involved.
A highly classified mission with extremely unclassified bloopers.
Forget donuts. We sent giant wooden horses to our competitors instead. Just a friendly gift… with absolutely nothing inside. Probably.
Review. Improve. Repeat. In this workshop, learn how to scale QA for AI agents.
Our 14.ai engineering dinner had everything: TypeScript, tzatziki, and team spirit. Shoutout to the brilliant guests who made it an unforgettable night.
We do onboarding a little differently here at 14.ai.
At 14.ai, every customer response kicks off parallel tasks: summarization, triage, urgency alerts. This snippet shows how we use Effect.Stream + Deferred to keep the system awake just long enough to finish the job.
When your product changes weekly but your KB doesn’t, AI agents go off-script. Here’s how to keep your support system sharp with a knowledge base that evolves at the speed of your team.
YC wasn’t just a launchpad: it was a group of people who cared deeply, moved fast, and made us feel like we belonged, even before we applied.
The year was 1776. The WiFi was “YeOldeGuest.” The founders signed, some in 72pt cursive.
Squeaky, versatile, occasionally soggy. A deep dive into Riva’s baguette-driven support architecture.
Learn how to build AI-powered knowledge bases that actually stay up to date, without rewriting the same article 47 times.
We hosted a Breakthrough Bistro dinner with special guest Colin Sidoti from Clerk.com.
Here's why the future of AI products depends on mastering the art of context engineering.
A hands-on guide to MCP, integrations, and designing agents that do more than just reply.
Raw fish, raw prompts, and stories from the front lines of AI support.
Pinkie is a scientist, food-tech founder, and the reason our team lunches feel like wellness retreats
This pen? Already sold. In a world of agentic systems, the pitch doesn’t matter: just let the AI close.
What if the Apollo 11 moon landing happened on Slack?
Everyone’s got a style: quadrants, speed runs, pancake art.
We hosted our first YC Founders Breakthrough Dinner with Bryan Onel of Oneleet and a brilliant crew of builders. From outbound strategy to pivoting fast, here’s what we learned (and why we’ll be doing more of these).
We’re diving into MCP and integrations, showing you how to make your AI support agents actually do things. Ideal for support teams, ops leaders, and AI builders looking to connect models to real-world workflows.
We celebrated a big milestone with an even bigger Slack-inspired celebration.
This recap guide from our workshop walks through how to design scalable, reliable AI support workflows using deterministic, non-agentic, and agentic patterns.
Vibe coding works, until it doesn’t. Here’s why AI needs tight loops, not blind trust.
Sometimes the smartest move… is the simplest one. Support efficiency, distilled.
Caught mid-focus. Please refrain from pings unless you're bringing snacks.
We kicked off our very first 14.ai event with sushi bowls, spirited conversations about agentic systems, and plenty of laughs (plus some love from Riva, the office dog). Just the beginning.
The slide everyone asked for: Planning-based agent design
This is your practical guide to writing production-grade prompts—from debugging system prompts and designing tool calls to using meta-prompts for scalable, reliable AI behavior in support workflows.
Writing our book in a Parisian café. Startup ideas, espresso, and the occasional waiter cameo.
At 14.ai, we believe prompting is the new interface—for using software and for building it. It’s a powerful shift in how teams create, automate, and define behavior in real time.
Reflections from our Paris offsite at Notre-Dame: on what holds things up, even when it’s not seen.
A glimpse into what customer support might have looked like during pyramid construction. Beer rations, camel APIs, and one very pointy top.
Why we didn’t just add AI to existing tools: we rebuilt support from scratch, agent-first.
Our YC interview lasted ten minutes. No script, no prep, just a Zoom call that kicked off a once-in-a-lifetime journey with Tom Blomfield and Tyler Bosmeny.
In this 30-minute advanced workshop, we'll go beyond triage and summaries to explore how to design prompts for real-world support agents, covering complex workflows, edge cases, and hardened prompt templates with built-in fallbacks.
At 14.ai, support isn't about tiers or ticket counts: it's about shared memory, fast coordination, and routing questions to the sharpest context. We're building that mindset into the product, too.
We rebranded to 14.ai. It's not a lucky number or a jersey: it's a mirror. A name that reflects our philosophy: no fluff, no faces, just digital intelligence built as it should be.
A hands-on guide to writing great prompts for your AI customer support agents.
No units. Just vibes.
We're excited to welcome Corey Mostero to the 14.ai team as our founding engineering intern. A Berkeley Applied Math student and Cal Hacks tech lead, Corey has been pushing to production since day one.
This week, founding engineer Maya Czeneszew attended the AGI House x Adaline Applied AI Summit, a private gathering of founders, CEOs, and product leaders focused on bridging the gap between AI capabilities and real-world product impact.
The Boston Tea Party as a customer support crisis: when tickets pile up, empires churn, and taxation needs representation.
We're hosting a workshop on building AI agents for support. This is a hands-on workshop where you'll learn how to write great prompts for your AI agents.
This week, our co-founder and CTO Michael Fester gave a talk at the AI Engineer World's Fair (June 3–5), a conference bringing together leading engineers and builders working on agentic systems, developer platforms, and AI infrastructure.
Introducing 14.ai: the end-to-end customer support platform designed from the ground up for AI-native teams.