Liquid AI is building tools for small-model and edge AI, and is looking for a hands-on Developer Relations professional to act as Liquid’s technical voice in the developer ecosystem. This hybrid role in San Francisco focuses on showing up where builders already work, shipping practical reference assets, and turning real developer feedback into product conversations across model and platform teams.
What you’ll do
- Be the technical voice in developer communities by engaging on Hugging Face, GitHub, Discord, and X, answering hard questions in public, maintaining Liquid’s presence on the Hub, and highlighting community fine-tunes and edge deployments.
- Host and represent in person by running hackathons, build nights, and technical workshops at Liquid offices and partner venues, and by submitting talks, running booths, and demoing in the hallway track at conferences and developer events relevant to small-model and edge AI.
- Build an on-device and adaptation story by shipping reference apps and technical resources that demonstrate LFMs on real edge hardware including iOS, Android, browser, Jetson, and NPUs.
- Maintain integration recipes for inference stacks developers use, including llama.cpp, MLX, ONNX, executorch, and the LEAP SDK.
- Publish practical cookbooks that help developers go from base model to fine-tuned, quantized, and deployed in a single sitting.
- Earn trust through technical content by writing deeply technical blog posts, demos, and honest benchmarks, including comparisons against Qwen, Gemma, Phi, and SmolLM, then converting strong workshop material into evergreen resources.
- Close the feedback loop by maintaining a friction log of developer pain points such as chat templates, tokenizers, and deployment gotchas, and bringing that signal into roadmap discussions with model and platform teams.
Core requirements
- Proven technical expertise with LLMs, including hands-on experience with model fine-tuning (LoRA, QLoRA, full fine-tuning, distillation), tokenizer debugging, and a track record of shipping to production or active community environments.
- Fluency with the modern AI stack, including PyTorch, Hugging Face (Transformers, PEFT, Datasets), and model serving frameworks such as llama.cpp, MLX, vLLM, ONNX Runtime, or TGI, plus an understanding of quantization tradeoffs like GGUF, AWQ, GPTQ, and INT8/INT4.
- Efficient model specialization, with experience in on-device deployment (iOS, Android, embedded) or specialized work within the efficient-model ecosystem (Phi, Gemma, Qwen, SmolLM, or distilled architectures).
- Developer relations and advocacy experience, including direct responsibility for engaging and supporting developers publicly.
- A strong public technical communication portfolio showing accuracy, clarity, judgment, and an authentic voice across technical articles, tutorials, cookbooks, talks, videos, workshops, or open-source educational content.
- Strong technical writing and editing skills to translate complex model, tooling, and deployment concepts into actionable content.
Technologies
PyTorch, Hugging Face (Transformers), PEFT, Datasets, llama.cpp, MLX, vLLM, ONNX Runtime, TGI, ONNX, executorch, LEAP SDK, GGUF, AWQ, GPTQ, INT8/INT4, LoRA, QLoRA, iOS, Android, Jetson, NPUs, browser, Discord, GitHub, Hugging Face.
Benefits
- Competitive base salary with equity in a unicorn-stage company
- Health: Liquid AI pays 100% of medical, dental, and vision premiums for employees and dependents
- 401(k) matching up to 4% of base pay
- Unlimited PTO plus company-wide Refill Days throughout the year
Nice-to-have
- Active open-source contributions, especially to inference or efficient-model tooling ecosystem (llama.cpp, MLX, ONNX, Hugging Face libraries)
- Experience organizing or running hackathons, workshops, or developer events
What success looks like in year one
- The LFM quickstart on Hugging Face becomes the cleanest path to production in the small-model ecosystem, with measurable lift in fine-tunes, downloads, and downstream Spaces.
- A library of on-device reference apps ships, with at least one each for iOS, Android, and an embedded target, supported by external developer forks, extensions, and releases.
- A consistent cadence of in-person events (LFM hackathons, edge-AI build nights, technical workshops) leads to measurable community growth.
- A working developer-friction feedback loop is in place: pain points are surfaced and fixed, OSS contributions land, talks get accepted, and inbound developer interest increases.
