DeepSeek Seeks 150 Engineers Amid Soaring AI Demand
Chinese artificial intelligence start-up DeepSeek plans to recruit approximately 150 senior back end engineers. This move addresses severe strain on its systems due to a significant increase in user demand and the rise of compute-intensive artificial intelligence agents. The Hangzhou-based organisation faces extensive infrastructure challenges.
Cui Tianyi, who joined DeepSeek in March to head its Harness team, revealed the recruitment drive. Mr Tianyi, a former quantitative trading expert, stated via social media that the company's existing back end infrastructure was approaching its operational limits. He wrote that scaling quantities in computing leads to a substantial increase in complexity. Upgrades, ongoing maintenance, and rewrites are necessary to support new technical directions.
And this focus on infrastructure engineering indicates a strategic shift for DeepSeek. The open-weight organisation is moving from a research-oriented entity to a high-volume platform. This development requires a stronger underlying technical foundation. The bolstered engineering capacity aims to deliver dependable, commercial services suitable for enterprise clients.
The current expansion follows previous growth initiatives. DeepSeek announced plans in June to significantly increase staffing, intending to "at least double the size of every department". That initiative involved opening 33 distinct positions across its research, engineering, and product management divisions. This points to sustained organisational growth.
But financial activities are also prominent for the AI firm. DeepSeek recently adjusted its Application Programming Interface access prices upwards, amid ongoing competitive pricing pressures among domestic AI companies. The organisation is also nearing the conclusion of a funding round.
This funding round places the company's value at approximately 500 billion yuan, or around US$74.5 billion, before new investment. According to a South China Morning Post report last month, citing informed sources, a potential listing on Shanghai's Star Market is under consideration. This financial activity provides context for the current operational expansion.
So operational demands remain high. The growing prevalence of autonomous AI agents contributes to these pressures. Unlike simple chatbot enquiries, a single agentic task often triggers multiple model calls. Each task also necessitates its own isolated computing environment.
This operational model requires DeepSeek to maintain substantial capital expenditure. Funds are directed towards servers, networking equipment, and software. The company continues this investment while working to keep its API pricing competitive against Western rivals.
And the 150 new back end dev roles specifically address these technical bottlenecks. These positions are divided across two primary technical areas. The first concentrates on server-side development. This encompasses large-model research platforms, agent frameworks, internal research and development infrastructure, public API maintenance, online services, and data engineering.
The second technical area focuses on elastic computing infrastructure. This involves both agent platform maintenance and lower-level system optimisation. The objective is to maintain fluid, responsive, dynamic compute workloads.
But the hiring announcement coincided with another significant company disclosure. DeepSeek confirmed it had begun beta testing for a new DeepSeek V4.1 Flash model. This latest version follows the original V4 series, first shown to the public in April.
A notice distributed within a WeChat user group provided details about the V4.1 Flash. The company stated that this model features a new architecture and includes native support for multimodality. DeepSeek indicated the model is "stronger in capability, and faster and cheaper," urging user testing and feedback.
DeepSeek will hire 150 senior back end engineers to address system strain.
The demand surge is attributed to increased user engagement and compute-heavy AI agents.
The company aims to bolster its infrastructure for enterprise-grade commercial services.
DeepSeek holds an approximate valuation of US$74.5 billion and explores a Shanghai Star Market listing.
Beta testing for the DeepSeek V4.1 Flash model, featuring multimodality, has commenced.
Source: SCMP
