Jobs at Embedding VC

• Brand-new function, real mandate — enterprise and team customers are one of our biggest growth levers, and most of the infrastructure behind them hasn't been built yet. You'll be in on the ground floor.
• Small team, massive surface area — engineers own real systems end to end, not slices.

• Ground floor of a new category — help define what "agent" means for creative work, the way Director redefined what "directing" means for video.
• Small team, massive surface area — engineers own real systems end to end, not slices.

• Ground floor of a new category — help define what "agent" means for creative work, the way Director redefined what "directing" means for video. You'll set the architecture everything else is built on.
• Own the whole agent layer — not a service, not a slice. Harness, orchestration, memory, evals, ...

• Brand-new function, real mandate — enterprise and team customers are one of our biggest growth levers, and most of the infrastructure behind them hasn't been built yet. You'll be the first engineer on it.
• Small team, massive surface area — senior engineers own whole systems, not slices.

• Brand-new function, real mandate — enterprise and team customers are one of our biggest growth levers, and most of the infrastructure behind them hasn't been built yet. You'll be the first engineer on it.
• Small team, massive surface area — senior engineers own whole systems, not slices.

• Design the end-to-end UX for OpenArt products—from mapping user journeys to crafting clear and engaging interactions.
• Act as the bridge between product and design, translating user needs and product requirements into elegant flows and experiences.

• Join the AI revolution! You'll bring AI-native tools into big enterprises.
• High ownership, speed, and accountability. We are a fast-moving, nimble team.

• Build vertical agents. Own end-to-end agents for high-value workflows across research, software, hardware, data, and operations—from problem definition through production use.
• Establish the agent foundation. Build reusable primitives for models, tools, orchestration, context, memory, retrieval, permissions, human approval, and long-running execution.

• Support the GTM Lead and founders across sales, growth, and content.
• Create technical content that makes a complex AI product legible to buyers.

• Build the GTM playbook from zero — positioning, target accounts, outbound motion, sales process, pricing inputs.
• Own enterprise sales end-to-end: prospecting, discovery, pilots, and closing.

• Build and ship product features across the stack on our new AI agent product.
• Prototype fast, learn from pilot customers, and iterate.

• Build the core of a new AI agent product end-to-end — backend and full-stack, from prototype to pilot.
• Turn rough product direction into shipped features fast, iterating with real pilot customers.

• Design, build, and ship full-stack product features — UI, API, and data layer — with a strong bias toward shipping and iterating quickly.
• Own features end-to-end, from product discussion through implementation, testing, and release.

• Own the architecture of Abaka Expert Platform's shared backend systems—core services, data models, APIs, and cross-product infrastructure.
• Define and drive the long-term technical strategy for backend architecture across the organization.

• Ride an inflection point. OpenArt has grown 7–10X in revenue over the past 2 years in one of the hottest categories in tech (AI-native creative tools). You're joining while the trajectory is still steep and every measurement decision compounds.
• Own the data science practice behind one of OpenArt's biggest growth levers — marketing spend a...

About Us

We are a small team obsessed with deep thinking, speed, craftsmanship, and a customer-first approach. Our mission is to build the best care experience in healthcare with a 30-year horizon.

About Us

Embedding VC is an early-stage venture fund investing at the frontier of AI. Embedding Labs is our sister studio, where we incubate and build AI-native companies from scratch alongside founders.

我们在训练自研视频生成基础模型(DiT / Flow Matching),需要一位既能搭起训练平台、又能把研究代码变成数百卡集群上稳定结果的工程师。你不只是用平台的人,更是建平台的人。

你会做

• 训练平台搭建:从作业调度、断点续训、监控告警到数据 / 权重流水线,把分散的脚本沉淀为团队可复用的训练基础设施。

• 数百卡规模的分布式训练:FSDP、张量并行、Context Parallel、Ulysses,把 MFU 推到合理水位。

• PB 级视频数据 pipeline:NVDEC 解码、VAE latent 缓存、变分辨率 bucket sampling。

• 显存与性能:...

岗位职责

视频标注

• 设计并落地各类视频标注任务与规范:caption(prompt 与视频内容对齐)、动作(action)、场景(scene)、风格、镜头 / 运镜、时序一致性、ranking / preference、badcase 分类等。

• 搭建标注质检体系:gold set、抽检、多轮 review、一致性(IAA)、reviewer 校准,保障大批量标注质量稳定可迭代。

• 管理标注供应商 / 众包团队,控制交付质量、一致性与返工。

• 根据模型 badcase 持续迭代标注规范(guideline)与标准。

视频评测

• 搭建视频生成评测体系:评测集 / ...

We help physicians to achieve independence by making it easier to start, operate, and grow their practices. We believe that the future of healthcare will not come from bureaucratic health care systems, but individual physicians.