# Solerno

> Solerno builds AI systems you can evaluate: environments, agents, and machines. Amsterdam.

Solerno B.V. is an applied AI lab in Amsterdam. We build environments, evaluation harnesses, agents, and machines — Systems you can score.

## Site

- [Solerno](https://solerno.ai/): Lab homepage
- [About](https://solerno.ai/about/): Who we are and how the lab works
- [Work](https://solerno.ai/work/): Named systems, not a services menu
- [Notes](https://solerno.ai/notes/): Essays that are not a product

## Work

- [FigmaBench](https://solerno.ai/work/figmabench/): A 100-task benchmark for whether language models can design in Figma through the Plugin API. Not screenshot-to-code.
- [Sudoku SFT + RL Environment](https://solerno.ai/work/sudoku/): A verifiable Sudoku environment for supervised fine-tuning and reinforcement learning. Constraints you can score, not vibes.
- [LangSmith Evaluation System](https://solerno.ai/work/evals/): How Solerno measures agents in production: traces, datasets, judges, and a loop back into the next run.
- [DeepTeam Red-Teaming with Local Qwen](https://solerno.ai/work/red-team/): Adversarial testing against a local uncensored Qwen 3.8 model. Safety work that does not depend on a hosted API’s filters.
- [Agent Ontology](https://solerno.ai/work/ontology/): A shared vocabulary for agents, tools, memory, and environments, so systems can compose instead of colliding.
- [Pi + Orca Local Environment](https://solerno.ai/work/local-dev/): The lab’s local development environment: agent orchestration, worktrees, and a machine you can actually see.
- [SO-101 Robot Arm](https://solerno.ai/work/so-101/): A tabletop arm on NVIDIA Orin Nano 8GB, ROS 2, and Gemini Flash. Physical intelligence at lab scale, not a robotics company claim.
- [İnanç Tekstil Curtain Store](https://solerno.ai/work/inanc-tekstil/): A live made-to-measure curtain shop: pick the fabric, set width and pleat, then try it in a photo of your own room before you buy.
- [Agentic Due Diligence](https://solerno.ai/work/due-diligence/): Long-running analysis over heterogeneous data sources. Agents that read, wait, and write a file, not a chatbot with a search bar.

## Notes

- [Primordial Field](https://solerno.ai/notes/primordial-field/): A philosophy note on fields, emergence, and what we think we are training when we train a model.

## Optional

- [Privacy Policy](https://solerno.ai/privacy/): How Solerno B.V. handles personal data
- [Terms of Service](https://solerno.ai/terms/): Terms for this website and lab work
- [Support](https://solerno.ai/support/): How to reach the lab
- [Full machine-readable index](https://solerno.ai/llms-full.txt): Longer summaries of every public page