Nitish Kulkarni

Today, we are launching NextToken - a single place to build production-grade agents, apps and analytics. Code-forward. Cloud Hosted. Zero setup. Extremely affordable. @nexttoken_co

@Jason I'm in

We started an AI founder twitter group... reply with "I'm in" if you're a founder and want to be added

@cognition Quite impressive! Are these triggered only when user explicitly asks or does the main session agent autonomously decide when to spawn off agents with separate sessions (as opposed to subagents in the same session)?

Devin can manage a team of Devins. Managed Devins run in parallel help break down complex tasks. Each managed session is a full Devin, with its own VM, terminal, browser, and testing infrastructure. The main session coordinates, monitors, and compiles results.

@maninblack815 @alankarjain91 Thanks Mihir!

Super excited to see where this leads! Congratulations @nitishkulki @alankarjain91!

Building useful real-world agents often comes down to system design: the right multi-agent architecture, shared context management, and carefully designed prompts + tools that let complex tasks run in parallel. Here’s one example of that in action: @nexttoken_co builds a multi-agent system to perform deep web research for influencer marketing - a common but highly complex task. Because Search + AI are built in, there are no extra API keys or integrations needed. But the real magic happens when instead of just dumping the agent's output into a file or table, @nexttoken_co turns it into an interactive app you can instantly use, share or host. All from a single prompt. 👇

Build real apps on top of powerful agents at nexttoken.co x.com/nitishkulki/st…

NextToken gives agents secure access to integrations. And it is fully code-first - not boxed into prebuilt tools, rigid templates, or fixed workflows.

Agent building should be more than “run” or “schedule.” You should be able to: - talk to your agents - inspect their sessions - see the steps they took - build real products on top

@claudeai Prompt to agents is the future! You can build managed agents for any model (not just claude) on nexttoken.co In addition to workflows, you can also build a frontend to talk to the agent!!

Introducing Claude Managed Agents: everything you need to build and deploy agents at scale. It pairs an agent harness tuned for performance with production infrastructure, so you can go from prototype to launch in days. Now in public beta on the Claude Platform.

Try NextToken: nexttoken.co x.com/nitishkulki/st…

The best part is what you can do next: Turn it into a dashboard, an alerting pipeline, or an agent that keeps monitoring the space over time. No context loss. No juggling multiple tools.

Creating a large web dataset can now be incredibly fast. Here’s one workflow we ran with NextToken: - find all mid-market companies for a criteria - enrich and analyze them - classify which ones have the highest AI automation ROI - draft tailored outreach emails All for under $1.

@prukalpa Thank you, @prukalpa !

Congrats on the launch @nitishkulki!

@sohamg121 Thank you, Soham!

I'm so excited for my friends to be building something so cool from scratch!

@im_mansigupta @alankarjain91 Thanks Mansi!

Something cool built by dear friends of mine. Check it out! @nitishkulki @alankarjain91

@im_mansigupta @alankarjain91 Thanks, Mansi!

@nitishkulki Many congratulations @nitishkulki and @alankarjain91 on the launch! The demos are quite impressive!

Today, we are launching NextToken - a single place to build production-grade agents, apps and analytics. Code-forward. Cloud Hosted. Zero setup. Extremely affordable. @nexttoken_co

@afrozenator @alankarjain91 Thanks, Afroz! :)

Many congratulations @nitishkulki and @alankarjain91 -- This is awesome and can't wait to see how this is used in the wild!!

If your work involves data, agents, or apps, we’d love for you to try it out: nexttoken.co We can’t wait to see what you’ll build. ❤️ x.com/nitishkulki/st…

Long-running agents fall apart fast when quality drops or costs explode. NextToken is optimized to run complex workflows reliably, at a fraction of the cost.