German engineering consultancy runs a self-hosted Qwen3 AI platform on four NVIDIA L40S GPUs for 200 employees
Particula Tech built a fully self-hosted AI platform for a 200-employee German structural engineering consultancy: Qwen3 models served with vLLM on a single server with four NVIDIA L40S GPUs, Open WebUI with Keycloak SSO, local RAG over project archives, Whisper transcription and a code assistant. Motivated by client NDAs and German data protection rules; the source reports replacing EUR 14,000/month of cloud AI spend with EUR 48,000 of hardware.
Overview
Particula Tech built a fully self-hosted AI platform for a 200-employee German structural engineering consultancy: Qwen3 models served with vLLM on a single server with four NVIDIA L40S GPUs, Open WebUI with Keycloak SSO, local RAG over project archives, Whisper transcription and a code assistant. Motivated by client NDAs and German data protection rules; the source reports replacing EUR 14,000/month of cloud AI spend with EUR 48,000 of hardware.
This entry has 16 published fields tied to exact passages in an immutable source capture.
Inspect the highlighted sourceThe challenge
About 80 employees had personal ChatGPT accounts they were pasting client data into, while monthly cloud AI spending had reached EUR 14,000 with no central oversight of what data was leaving the network. Client NDAs on most infrastructure projects prohibited sending project data to third-party cloud services, and German data protection regulations added a further constraint.
The solution
Particula Tech built a self-hosted platform on a single server with four NVIDIA L40S GPUs running open-source Qwen3 models served with vLLM, with Open WebUI as the company-wide interface behind Keycloak SSO, a local RAG knowledge base in Qdrant over the firm's project archives and the DIN and Eurocode standards, local Whisper transcription with meeting summaries, and a Qwen2.5-Coder-32B code assistant.
Reported business value
The source reports the platform eliminated EUR 14,000 per month in cloud AI subscriptions and API costs and, with EUR 48,000 of hardware and about EUR 400/month in electricity and maintenance, paid for itself in 3.5 months. It reports adoption of 94% of employees by month three. The firm's estimate of about 15 hours saved per engineer per month on research is described by the source as self-reported and likely optimistic.
Sources
Open any source and check the claim yourself — that is the point of the register.
This record was researched and written with AI assistance, and its claims were checked against the sources above. (EU AI Act art. 50 transparency notice.)
Other professional services entries in the register.
McKinsey's Firmwide Knowledge Copilot "Lilli"
McKinsey launched Lilli, a generative-AI platform that provides streamlined, impartial search and synthesis of the firm's internal knowledge to help consultants build client engagement plans. The firm's intellectual property spans more than 40 curated knowledge sources today, and McKinsey said there will be more than 100,000 documents and interview transcripts of internal and third-party content in the future. A user asks a question and Lilli scans the firm's knowledge base, surfaces five to seven of the most relevant pieces of content, summarizes key points, and identifies subject-matter experts. McKinsey piloted Lilli with real users and was scaling it from an initial team of three to more than 70 experts, aiming to reach thousands of colleagues by the end of the year.
PwC uses AI-driven send-time and audience scoring to improve campaign relevance
PwC France & Maghreb embedded AI models in Oracle Marketing to determine optimal send times, detect audience fatigue to avoid over-contacting, and score accounts by engagement potential and service-line interest, shifting from broad regional campaign lists to intent-based segments.
Prosegur and Telefónica roll out the Gen IA generative AI agent for internal IT support for more than 25,000 users
Per Euro Security (8 Jan 2026), Prosegur worked with Telefónica to roll out the generative AI agent Gen IA across its European and Latin American branches, taking over central management of internal IT support for more than 25,000 users via the group-wide service desk, built on Microsoft Azure services. Prosegur reports a Net Promoter Score of nearly 75 per cent, up to 88 per cent of enquiries handled fully autonomously, and a 16 per cent reduction in enquiries previously handled by phone support staff and technicians. A worldwide rollout to around 180,000 employees is planned.
ADT embraces Oracle AI to improve onboarding and enhance workforce insights
ADT, the smart home security provider that hires more than 5,000 new employees annually, deployed Oracle Fusion Cloud HCM's embedded AI features, including the Onboarding Assistant AI agent, which gives new hires guided, step-by-step, 24/7 support through onboarding, and author-assistance AI for performance management and goal setting. Oracle Fusion HCM Analytics uses AI to analyze data from HR transactions, help desk tickets, performance check-ins, and self-service activity to surface attrition patterns, performance trends, and manager effectiveness. ADT reports managers are saving two to three hours per direct report on performance evaluations and goal setting, amounting to a full week regained per year in some cases, plus more consistent onboarding and reduced HR support ticket volume.
Was this helpful?
Your feedback helps us improve our use case database
