Corporate communications
News from the machine room. And four open seats in it.
Company news
June 2026 · Partnership
Localized AI joins the Anthropic partner network for Claude implementations
We are pleased to announce that Localized AI has been selected as part of Anthropic's Claude partner network. For customers whose use cases call for frontier-model capability, we now implement Claude models — combined with our signature architecture: sensitive document processing stays on-premises, while Claude handles orchestration and advanced reasoning. The same worksheet, one more excellent option in it.
Claude Partner Network · AnthropicMay 2026 · Engineering
World-model pilot: packaging line predicts its own jams
Our first world-model deployment is live at a manufacturing customer: a compact model trained on the line's own camera footage flags developing jams 1–2 seconds before they form. How it works →
April 2026 · Content
26 engineering guides published
Our complete guide library is live — fine-tuning, RAG, Azure AI, humanoid robotics, with working code in Python, C#, Go and Node.js. Read the blog →
Press contact: ai@localized-ai.com
Open positions
Small team, sharp tools, customers who say thank you. Every role ships things that run inside real companies.
Machine Learning Engineer — Fine-Tuning & Evaluation
You own the training pipeline: dataset curation with customer experts, LoRA/QLoRA fine-tuning, quantization, and the evaluation harness that decides what ships.
What we look for
- Solid Python and PyTorch; you have fine-tuned and shipped at least one transformer model
- Experience with PEFT, llama.cpp/Ollama or vLLM; quantization trade-offs are not news to you
- You build evals before demos and can explain a regression gate to a customer
- German B2+ — you will sit with domain experts reviewing training data
Infrastructure / DevOps Engineer — On-Prem AI Platforms
You make GPU servers boring: hardened Linux, sealed VPN deployments, monitoring, backup and the outbound-deny firewall proof our customers sign off on.
What we look for
- Strong Linux administration, networking (VLANs, VPN, firewalls) and Docker
- Experience operating GPU workloads; NVIDIA driver/CUDA pain tolerance acquired
- Ansible or similar — our deployments are checklists turned into code
- Willingness to occasionally stand in a customer server room in Hintertupfingen
Solutions Engineer DACH — Discovery to Deployment
You run discovery workshops, translate "we drown in tickets" into a scoped use case, and stay on the project until the eval gate is green and the customer is trained.
What we look for
- Technical background (software/data) plus the ability to talk to a Geschäftsführer and a works council in the same afternoon
- You can demo RAG, explain GDPR-by-architecture, and write an honest cost comparison
- German C1+ and English fluent
- Experience in B2B consulting or pre-sales engineering a strong plus
Working Student — AI Engineering (m/f/d)
Real work, not coffee: you build eval datasets, test model candidates, maintain our internal benchmark suite and write the occasional blog deep-dive with us.
What we look for
- Enrolled in CS, math or related; Python comfortable
- Curiosity about LLMs beyond using them — you want to know why the loss curve bends
- Reliable, self-organized, and honest about what you don't know yet