exosphere.host

@spar_raj PR here: github.com/exospherehost/… Available now in beta

PR Merged: TTL-based cleanup strategy for DatabaseTriggers MongoDB collection (Contributed by @spar_raj )🥳 Developers can now set a expiry timeout in env, all used records which meet the timeout criterion would be deleted automatically, reducing db pressure and storage cost.

Some evenings are special. Not because you’re out on a date, but because you get to showcase your product in front of 200+ brilliant engineers across domains. This Tuesday at @Magicball_dev Reveal, I had a blast talking about agents, infra, AI, and scaling challenges. We’re building @exospherehost to help developers reliably scale and parallelize AI agents and workflows to massive levels, with deep observability and a developer-first experience. If we missed each other at the event, or if you just want to jam on AI, infra, or scale, my DMs are always open.

Thank you @kalraa_v way to go🚀

See you in FOSS 2026!

Exosphere at IndiaFOSS 2025! Learn how to scale AI workflows with just a few lines of Python. Built in the open for the community. Scalability and reliability for every developer. @IndiaFOSS @FOSSUnited

@niveditjain lets scale 🚀🚀

Most AI workflows fail not because of the model, but because of the architecture. Introducing Exosphere: - Parallelization - Reliability - Developer Experience To scale any AI workflow. Link in comments.

PR Merged! ( Contributor @kalraa_v, reviewer: @niveditjain ) 🧑‍💻 This PR enables Auto-Refresh functionality will enable developers will enable developers to observe their workflows better and significantly improve the developer experience.

my first ever open source contribution got merged to @exospherehost 🌟

It was super fun to present @exospherehost yesterday at Agentic Summit BGLR by @magicball_dev. Apart from showcasing our tech, we had great fun chatting and soaking in insights. If we missed connecting during the event, do reach out. Happy to meet/emeet

@exospherehost the platform for async ai workflows Similar to JS event loop but for AI tasks Runs AI tasks as a background job, makes it really cheap, with the tradeoff of time. Really useful if you wanna just put all the tasks in the queue and sleep lol exosphere.host

🧁

🚀 Meet @exospherehost We just wrapped up a deep sprint last week: finalizing our architecture, detailing our first set of satellites, and laying down the core foundation. This week is all about building and showing you what we've been working on. Exosphere is the infra layer for large-scale AI workflows. Massive data, large models, multi-step jobs? We handle it. No cloud setup, no infra ops. Just connect components with a simple YAML or Python SDK and go live. Already being used to run long, resource-heavy agents and workflows in the background, resilient, cost-efficient, and scale-ready. DMs open if you're building in AI. We'd love to hear what you're working on. What kind of workflows or agents are you building? Drop a reply.

@niveditjain LFG 🚀

@niveditjain 🤯🚀

Join @exospherehost at the Agentic Summit in Bengaluru We building an open-source infrastructure designed for robust, cost-effective, and easy-to-use AI workflows and intelligent agents. We look forward to connecting with developers, creators, and innovators. Your feedback is highly valuable to us. Attaching a short preview video. Registration link below.

@niveditjain @dok2001 @accelindia @8vc @grayscale_vc @stripe @googlecloud Coooool🥳🚀

LFG 🚀 Thrilled to share that I would be showcasing Exosphere at the Agentic Summit BLR on July 2, 2025! Looking forward to show off what we are building and learn from developers, especially from @dok2001 ✨ Would be waiting at the @exospherehost booth to say hi! 👋

Looking forward meeting you 😉

@niveditjain Let's gooo🚀🚀🚀

I am super pumped to release a preview of exosphere cluster APIs! This will allow developers to ship custom AI workflows in matter of minutes, no setup, no infra pain, just a simple declarative YML file and an API call, yes that's all you need. And yes, up to 75% cheaper inference with exosphere!! How? checkout👇