New comment by Slate_official in "Ask HN: What are you working on? (March 2025)"

Slate_official — Fri, 04 Apr 2025 02:53:11 +0000

I recently made the leap from product management to coding, with a lot of help from GPT-4 (it basically taught me how to code). I’ve been obsessed with LLM workflows and kept running into the same pain: constant context resets, tool-switching, and token limits that kill flow mid-task.

So I built Slate — an AI-powered workspace that supports long-context reasoning (up to 1M tokens), persistent memory, and lets you work with multiple models (OpenAI, Claude, Gemini, etc.) in a Notion-like interface.

It’s still early, but live. If anyone here experiments with long-context LLM use cases or AI tooling, I’d love your feedback — and I’m currently giving away credits to try the Gemini models because I’ve got $300 worth that expire in 48 hours. No catch, just want it used and tested.

First time building and launching something on my own — excited but also a bit overwhelmed. Appreciate any thoughts!

Hacker News: Slate_official

New comment by Slate_official in "Ask HN: What are you working on? (March 2025)"