We are building the next generation of intelligent creative tools.

If you are among the best AI engineers working at the intersection of computer vision, storytelling, and multimodal AI, this is a rare opportunity to define how machines perceive and support narrative construction in visual media.

Why Join

You will work alongside leading AI and vision researchers and collaborate directly with Hollywood-level creatives and editors. Our team is uniquely positioned to blend cutting-edge machine learning with the art of story creation, shaping tools that elevate how humans craft visual stories—from idea to screen.

About Calli (stealth)

Calli is a stealth-stage AI venture rethinking how stories are discovered, structured, and enhanced through intelligent systems. We combine expertise in machine learning, software engineering, and visual storytelling, and are growing to 4 – 8 people spanning AI, full-stack, and creative fields.

What You Will Do

• Build and train advanced models for video and image understanding, person and object tracking, action recognition, and scene parsing.
• Contribute to systems that support visual narrative creation, identifying, structuring, and editing meaningful moments across video content.
• Develop and maintain end-to-end training pipelines on real-world datasets.
• Analyze and present results through visual analytics tailored to engineering and creative collaborators.
• Shape both the technical foundations and product features of our first release.

What You Bring

• 3+ years of hands-on experience training deep learning models from scratch.
• Proven ability to build and debug your own neural networks (not merely fine-tune).
• Deep proficiency in PyTorch and familiarity with Hugging Face tools and model hubs.
• Experience with some of the following (you need not know all):
• 
  
• Object detection & tracking – e.g. YOLOv5/v8, Detectron2, ByteTrack, or similar
• Person identification & re-identification – e.g. OSNet, DeepSORT, ArcFace, or similar
• Action recognition – e.g. TimeSformer, VideoMAE, SlowFast, or similar
• Multimodal / foundation models – e.g. CLIP, Flamingo, BLIP-2, DINOv2, or similar
• Experience scaling experiments with DDP, Lightning, Accelerate, mixed precision, etc.
• Skill in analyzing, visualizing, and communicating results.
• Good engineering hygiene: clean, modular code; Git; reproducible setups.
• Bonus: publications, impactful deployments, open-source contributions, or bridges from research to product.

Nice to Have

• Experience with generative models for vision (diffusion, GANs, etc.).
• Built or curated video-text / image-text datasets.
• Work on video summarization, scene segmentation, or story-structure inference.
• Familiarity with narrative theory or film-editing tools (helpful but not required).

What We Expect

• A two-page CV detailing academic and industry experience.
• A GitHub or portfolio featuring at least one deep learning project you trained end-to-end.
• Ability to work under uncertainty, iterate quickly, and communicate clearly.
• Availability for hybrid collaboration, with regular in-person meetings in the NL/EU region.

What We Offer

• A central role in building AI tools that elevate human storytelling.
• Close collaboration with world-leading experts in AI and narrative craft.
• Competitive salary (wide range based on experience) plus equity upside.
• Flexible hybrid setup: Netherlands preferred, anywhere in Europe considered.
• Fast and respectful hiring (two–three steps, including technical discussion).
• A culture of clarity, momentum, and trust.
• We welcome applicants from all backgrounds and identities.