Put sales on autopilot

gm legends, happy Monday.

Here’s today’s lineup: Oskar by Skarbe handles the follow-ups, lead chasing, and deal tracking you keep forgetting to do so you can actually focus on building; Lumi from Google’s PAIR team turns reading research papers into an interactive experience with AI that reads beside you; and Magic MCP by Metorial gives AI agents instant access to 600+ integrations with almost no setup required.

P.S. Building something new? Tell us about it → editorial@producthunt.co 🫶

Oskar by Skarbe is an AI agent that lives inside your inbox and quietly handles the stuff you never get around to, qualifying leads, following up, enriching contacts, and tracking deals while you’re off actually running the business.

🔥 Our Take: You know that sinking feeling when you realize you forgot to follow up with someone who was ready to buy? Yeah. This kills that. It doesn’t make sales “fun,” it just makes sure it happens, even when you’re buried in everything else.

It’s November, and with the holidays around the corner, it may not feel like hiring season—but it very much is, according to fmerian. He’s compiled his largest list yet of job vacancies. So whether you’re a developer relations engineer in San Francisco, a product engineer in London, a full stack developer in the EU, or something somewhere in between, start browsing.

Magic MCP by Metorial gives AI agents plug-and-play access to 600+ integrations through simple OAuth and serverless deployment. It’s open-source, fast to scale, and built on the Model Context Protocol so your agents can actually get things done instead of sitting in setup hell.

🔥 Our Take: Hooking up APIs is the kind of grunt work everyone pretends to automate but never really does. This finally does. You drop in the keys, flip the switch, and watch your agent go from sandbox toy to something useful.

Lumi is an AI reading experiment from Google’s PAIR team that adds an interactive layer to arXiv papers. It highlights key ideas in the original text, turns references into links, and lets you ask questions about any section or image all within the paper itself. Fully open-source under Apache 2.0.

🔥 Our Take: Academic papers are great at two things: burying insights and breaking focus. This tool finally gives you a fighting chance. It keeps you in the text, helps you follow your curiosity, and makes research feel a little less like punishment.