Skip to main content

DeepSeek's Big Hiring Push: 150 Engineers to Build the Next-Gen Agent Platform

DeepSeek is making waves again, but this time it's not about a new model release. On September 8, the company announced a massive recruitment drive, aiming to bring on board around 150 engineers. The focus? Server-side development and Agent elastic computing research. They're looking for senior backend engineers with 2 to 10 years of experience, and notably, they're not hiring for AI research positions at all.

This is a bold move, especially coming just two months after DeepSeek announced a full-department expansion back in June. It's clear that the company's strategy is shifting. Instead of just pushing the boundaries of pure model research, they're now doubling down on the engineering side—building the infrastructure that will make AI agents actually work in the real world.

So, what exactly are these engineers going to work on? The recruitment spans six server-side areas, including the large model research platform, Agent framework components, DeepSeek API, and online services that support tens of millions of daily active users. But the real meat is in the elastic computing platform roles. These positions demand engineers who can squeeze every last drop of performance out of the hardware, pushing the limits of system optimization.

This is all tied to DSec, DeepSeek's self-developed elastic computing platform, which was first revealed in the DeepSeek-V4 technical report. DSec is built on the 3FS distributed file system and is composed of three Rust components: API Server, Edge proxy, and Watcher. The company has big plans to overhaul the entire system stack—from the operating system and virtual machine down to network and storage. The goal is to ensure that Agent model training can happen efficiently and iteratively, creating a closed-loop production workflow that goes from execution to recording, feedback, and finally, nourishment.

But why all this effort? The AI talent market is brutal right now, and DeepSeek has recently faced the loss of some core R&D personnel. To counter that, they're leveraging their latest funding round of 5.1 billion yuan and a valuation of 340 billion yuan on the Hurun Unicorn List. With such deep pockets, they can offer highly competitive compensation to attract the most skilled engineering talent out there.

This move comes at a time when large model capabilities are becoming increasingly similar across the industry. The real differentiator now lies in the underlying systems—things like elastic scheduling and sandbox isolation. DeepSeek's hiring spree is a clear signal that the competition has shifted to the engineering battlefield. Whoever can build a SOTA-level Agent runtime environment first will have the power to define the next phase of industry standards.

In a nutshell, DeepSeek is betting big on infrastructure. They're not just looking for coders; they're looking for engineers who can build the backbone of the next generation of AI agents. And with this kind of investment, they're clearly aiming to lead the pack.

Key Points

  • DeepSeek is hiring 150 senior engineers, focusing on server-side development and Agent elastic computing.
  • The recruitment targets engineers with 2-10 years of experience, with no AI research positions.
  • This follows a full-department expansion announced in June, signaling a strategic shift towards engineering.
  • The engineers will work on DSec, DeepSeek's elastic computing platform, and other infrastructure.
  • DeepSeek is using its $5.1 billion funding to attract top talent in a competitive market.
  • The move highlights the industry's shift towards underlying system engineering for Agent implementation.