Tyler Haun

🚀 12 Startups that have announced a funding round this month that are hiring! Mastra → TypeScript framework for AI agents | $13M Seed | Oct 2025 Relace → Infrastructure for AI coding agents | $23M Series A | Oct 2025 Kernel → Browser infra for web agents | $22M Series A | Oct 2025 supermemory → Open-source AI memory platform | $3M Pre-Seed | Oct 2025 Reflection AI → Autonomous coding agents & superintelligence | $2B Series B | Oct 2025 Datacurve → AI-coding training data infrastructure | $15M Series A | Oct 2025 Spellbook → AI copilot for transactional lawyers | $50M Series B | Oct 2025 n8n → AI workflow automation tool | $180M Series C | Oct 2025 David AI → Data layer for audio AI | $50M Series B | Oct 2025 Base Power Company→ Building America’s next-gen power company | $1B Series C | Oct 2025 Stoke Space → Fully reusable rockets | $510M Series D | Oct 2025 Crosby → AI-enabled contract law firm | $20M Series A | Oct 2025 Will link careers pages in comments

20

1 Comment

Anthropic just dropped Claude Sonnet 4.6, and it’s officially time to rethink your AI strategy. For a long time, the rule of thumb was: "Sonnet for speed, Opus for the heavy lifting." That line just got incredibly blurry. Sonnet 4.6 isn't just an incremental update; it’s a full-scale upgrade that rivals the flagship Opus 4.6 while staying at the Sonnet price point. We are moving away from AI as a "copy-paste" assistant and into the era of Autonomous Workflows. Whether you are building complex agents or just trying to ship code faster, Sonnet 4.6 is the new gold standard for what a "mid-tier" model can do

3

"Context engineering” is just prompt engineering with added context. The LLM looks at two things: 1. Prompt = Guard rails, Format, Instructions, etc... 2. Context = What documents/code/chat history to include, Summary, etc... How is it different from plain prompt engineering? The queries are heavily augmented here, with chat history summary and other relevant documents. Think of it as RAG + structured prompts. Calling it "engineering" is a stretch. It's too broad and domain-specific to be a science. If you are a software engineer looking to learn more about AI engineering, try this structured course at InterviewReady: https://lnkd.in/gAFcCkRj Cheers! #ContextEngineering #LLMs #AI

55

5 Comments

Taking bets on when we will see Supabase -> Supabase -> 💰to Supabase ...? On a more serious note, a lot of people are rightly concerned about startups that are just thin wrappers around someone else’s AI model. At leap.new, we use powerful models from Anthropic and Google, but what really sets us apart is our integration with Encore (which we also built). Leap is the only AI agent that can build truly full-stack applications and deploy them directly to your own cloud infrastructure on AWS or GCP. We believe this is just the first step in our long term vision to enable AI to built large-scale software at Human+ capability.

52

6 Comments

If you're building with LLMs, watch out for this hidden cost... LLMs are evolving faster than your prompt stack can keep up. What breaks isn't always visible. Until it is. Every model shift carries silent regressions that chip away at velocity. The more LLM logic you ship, the more fragile your system becomes. Here's where the real cost shows up.

9

1 Comment