About Us
We build fast, modern developer tools in Rust, and a surprising amount of serious software runs on them.
Pixi, rattler, and rattler-build are open source and used by hundreds of thousands of developers: quant teams in finance, physicists at CERN, biotech labs, and humanoid robot companies you’ve probably seen videos of.
We are also core maintainers of conda-forge, one of the largest open source package ecosystems, and we maintain RoboStack, which gets ROS running on any OS from a single lockfile.
When that much software flows through your tools, security stops being optional. Every package that ends up on a laptop, a server, or a robot should be traceable back to the source and build that produced it.
We’re building that for conda today, and we want to extend it to AI models.
Our engineering team is spread across Europe (Netherlands, Germany, UK, and Moldova). This work is based in Berlin, and so is this role.
The Work
Robots increasingly run learned models: vision transformers, RL policies, and LLMs. Today, those models are usually copied onto machines as large, opaque files, with little record of where they came from or what changed.
We think models should be shipped like packages: versioned, signed, verifiable, and updated efficiently.
You’ll work on two related areas:
1. Package Security
- Signing and verification for conda packages with Sigstore.
- Provenance from source to build to install.
- Software Bill of Materials (SBOMs).
- Mapping known vulnerabilities onto the packages people actually use.
- Contributing this work to Pixi, rattler, and conda-forge, where it will be used by a large community.
2. AI Models as Packages
Taking the same ideas to model artifacts. That includes:
- Model provenance: Building a provenance chain from training through packaging to deployment, with signatures that can be verified offline on air-gapped robots.
- Model SBOM: Developing a model SBOM format, building on CycloneDX’s ML-BOM.
- Delta updates: Developing updates between model versions (full models, fine-tunes, LoRA adapters, and quantized variants), so robots don’t download gigabytes for a small change.
- Compression: Combining model techniques (quantization, pruning, and distillation) with what package managers already do well (zstd and content-addressable storage), tuned to the target hardware and network.
- On-device fine-tuning: Managing models that get fine-tuned on the robot itself, tracking where those variants came from, validating them, and syncing them back.
- End-to-end testing: Testing the whole pipeline on real models and real hardware, and measuring latency, bandwidth, and model quality.
A lot of this is new. Part of the job is figuring out which approaches work, measuring them honestly, and writing up what you found.
What We’re Looking For
You don’t need years of experience. You do need solid fundamentals, curiosity, and the patience to get security details right.
- Programming skills: Experience with Rust, or a strong background in C++, Go, or Python and the wish to switch.
- Software supply chain security: An interest in software supply chain security. Bonus if you’ve worked with Sigstore (Cosign, Rekor, Fulcio), TUF, in-toto, SLSA, or SBOM formats.
- Machine learning: Familiarity with ML in practice. You’ve trained or fine-tuned a model, know what a LoRA adapter or a quantized model is, and have used PyTorch or similar.
- Cryptography: Basic understanding of cryptography as an engineer, including signatures, hashes, certificates, and what they do and don’t guarantee.
- Analytical mindset: You like measuring things and are honest about results that don’t work out.
- Location: You live in Berlin or are willing to move here.
It’s a plus if you:
- Know package managers from the inside.
- Have worked on model compression or edge inference.
- Have robotics experience.
- Have contributed to open source projects.
None of these additional qualifications are required.
What We Offer
- Meaningful technical challenges: Work on problems few people have solved yet, together with the people who created Mamba, Pixi, rattler, and RoboStack.
- Open source impact: Most of your work will be open source and used by a large community from day one.
- Competitive compensation: We offer a competitive Berlin salary.
- 30 days of vacation: Plenty of time to relax and recharge.
- Hardware of your choice: Get the hardware you need to do your best work.
- Conference budget: Opportunities to attend conferences and continue learning.
- Open source contributions: Dedicated time for upstream open source work.
- Relocation support: We help with relocation within the EU.