高级AI工程师
Senior AI Engineer
Grip 正在将企业内容制作从手动、工具驱动的工作流程转变为可编程系统,实现大规模生成高质量视觉输出。作为高级 AI 工程师,你将构建推动这一转变的控制层:AI 控制器、可重用模板和工作流基础组件,将品牌意图转化为确定性的、可投入生产的输出。
这不是提示工程。这是设计系统,将生成模型约束、引导并扩展为全球品牌使用的可靠创意生产流程。
要求
你将负责的工作
构建 AI 控制的内容系统
- 设计并交付 AI 控制器,将视觉逻辑(构图、布局、约束)编码为可重用系统
- 将创意方向转化为结构化流程(例如,将标志安全区域转化为可执行的空间约束)
- 设计模板,在数千种变体中生成一致的输出
开发并扩展 AI 工作流
- 构建并优化 ComfyUI 流水线,整合多个模型、优化器和控制机制
- 结合分割、深度图和潜在约束,引导生成达到生产级输出
- 根据输出质量、故障模式和客户要求迭代工作流
负责模型行为和输出质量
- 与扩散模型、编码器和潜在空间操作进行实际操作
- 定义提示、条件和控制信号在流水线中的交互方式
- 确保输出符合品牌和视觉标准——不仅是“好的图像”,而是可用资产
扩展平台功能
- 为 Grip 平台中的新节点定义和 AI 能力做出贡献
- 在 Python/PyTorch 系统中为新功能指定技术需求
- 与工程师和创意团队合作,推动自动化内容生成的边界
你是什么样的人
你构建的系统让 AI 可预测、可控且可投入生产——尤其是在视觉领域。
你具备:
- 在生成式 AI 系统方面的丰富经验(扩散模型、条件生成、潜在空间控制)
- 在 ComfyUI 或类似基于节点的流水线中有实际操作经验
- 扎实的编程能力(Python、PyTorch),有交付运行系统的经验
- 有训练或微调模型以及管理其生命周期的经验
- 接触过 Photoshop、Blender 或类似视觉工具(不是思考——是实际使用)
你思考的方式:
- 以系统思维,而非提示——关注组件之间的交互、失败和扩展性
查看英文原文
Grip is shifting enterprise content production from manual, tool-driven workflows to programmable systems that generate high-quality visual output at scale. As a Senior AI Engineer, you build the control layer behind that shift: AI Controllers, reusable templates, and workflow primitives that translate brand intent into deterministic, production-ready outputs.
This is not prompt engineering. This is designing systems that constrain, guide, and scale generative models into reliable creative production pipelines used by global brands.
Requirements
What You Will Do
Build AI-Controlled Content Systems
- Architect and ship AI Controllers that encode visual logic (composition, layout, constraints) into reusable systems
- Translate creative direction into structured pipelines (e.g., turning logo safe zones into enforceable spatial constraints)
- Design templates that produce consistent outputs across thousands of variations
Develop and Scale AI Workflows
- Build and optimize ComfyUI pipelines integrating multiple models, refiners, and control mechanisms
- Combine segmentation, depth maps, and latent constraints to guide generation toward production-grade outputs
- Iterate on workflows based on output quality, failure modes, and client requirements
Own Model Behavior and Output Quality
- Work hands-on with diffusion models, encoders, and latent space manipulation
- Define how prompts, conditioning, and control signals interact across the pipeline
- Ensure outputs meet brand and visual standards—not just “good images,” but usable assets
Extend Platform Capabilities
- Contribute to new node definitions and AI capabilities within Grip’s platform
- Specify technical requirements for new features across Python/PyTorch systems
- Collaborate with engineers and creative teams to push what’s possible in automated content generation
Who You Are
You build systems that make AI predictable, controllable, and production-ready—especially in visual domains.
You Have:
- Strong experience with generative AI systems (diffusion models, conditioning, latent space control)
- Hands-on work in ComfyUI or similar node-based pipelines
- Solid programming skills (Python, PyTorch) with experience shipping working systems
- Experience training or fine-tuning models and managing their lifecycle
- Exposure to visual tools like Photoshop, Blender, or similar (not optional thinking—actual usage)
You Think:
- In systems, not prompts—how components interact, fail, and scale
- In constraints—how to guide models toward predictable outputs
- In outputs—what makes something production-ready vs. visually interesting
- Pragmatically—you ship working solutions and refine based on results
Your Skills Are:
- Translating abstract creative direction into structured AI workflows
- Debugging generation pipelines across multiple control layers
- Balancing flexibility with control in template systems
- Communicating across engineering and creative teams
Why Join Grip?
The Platform Position.
Grip is infrastructure for enterprise creative production—powering global brands to generate content at scale with control and consistency.
The Market Timing.
AI-generated content is moving from experimentation to production. The challenge is no longer generation—it’s control, reliability, and scale.
The Technical Moat.
You’ll work with ComfyUI-based workflows, multi-tenant architecture, and a stack spanning TypeScript, Python, Elixir, and C#—built for production, not demos.
The Work.
You will define how AI is actually used in real-world creative systems—turning unpredictable models into dependable pipelines.
The procedure
If you’re ready to bring your powers to a role where you can make a difference, we want to hear from you. Send your resume and portfolio in English via the link provided. If you have questions, contact our Recruitment Department at .
INDG Grip handles and uses personal data of job applicants in line with its Recruitment Privacy Policy.
Originally posted on Himalayas