Description

Member of Technical Staff (Research Engineer) embedded in production training of Black Forest Labs' large multimodal generative models. Improves the performance, reliability, and numerical stability of training runs across large GPU fleets, profiling full training steps and implementing GPU-level optimizations like fused kernels and low-precision paths. Debugs distributed training failures such as NaNs, loss spikes, and NCCL issues, and builds benchmarking harnesses that make performance claims trustworthy.

Requirements

  • Deep experience with large-scale training systems alongside researchers
  • Strong PyTorch fluency, including low-level training code
  • Experience with FSDP, tensor/sequence parallelism, NCCL, and checkpointing
  • Hands-on improving training throughput, memory, or stability
  • Experience profiling GPU workloads with Nsight or torch profiler
  • Understanding of low-precision training and quantization tradeoffs

Experience Level

InternEntry LevelMid-levelSeniorLeadExecutive

Benefits

Equity

💰 Net Salary Calculator

Tax class 1, no church tax
Gross Salary
€185,000/year
€15,416/month
Total Deductions
€80,657
44% effective rate
Net Salary (Take Home)
€8,695/mo
€104,343/year

Deduction Breakdown

🏛️ Income Tax€58,893
🏦 Pension Insurance€9,430
🏥 Health Insurance€6,103
🫶 Care Insurance€1,674
📋 Unemployment Insurance€1,318
⚖️ Solidarity Surcharge€3,239

🏠 Cost of Living in San Francisco

Monthly Essentials

Rent (1BR)
€3400
Groceries
€570
Transport
€86
Utilities
€256
Internet
€74

Lifestyle & Entertainment

Eating Out
€27
Gym/Month
€61
Coffee
€7
Beer (0.5L)
€11
Cinema
€20
Monthly Net Salary:€8,695
Comfortable Living:€4,416
Monthly Savings:€4,279 (49%)
💰

Strong savings potential: You could save €4,279 monthly, which is €51,348 per year.

💡 These estimates are based on average costs in San Francisco. Actual expenses vary based on lifestyle, neighborhood, and personal choices. "Comfortable living" includes rent, utilities, groceries, transport, and reasonable dining/entertainment.

📍 Life in San Francisco

👥
Population
0.8M
⭐
Quality of Life
8.0
🇬🇧
English Friendly
10.0
💻
Tech Scene
10.0
🌍
International
9.0
🚇
Public Transport
8.0
🚴
Bikeability
7.0
🌬️
Air Quality
8.0
🌳
Green Spaces
8.0
🚗
Car Free
6.0
ℹ️

About San Francisco

  • •OpenAI, Anthropic, Salesforce, Stripe, Airbnb, Uber and many AI startups
  • •Densest concentration of venture capital in the world
  • •Very high salaries and equity compensation
🌤️

Climate

  • •Cool Mediterranean climate
  • •Foggy summers
  • •Mild, rainy winters
❄️Jan: 11°C☀️Jul: 17°C📅Avg: 15°C
☀️3060 sunshine hours/year
🛡️

Safety

  • •Property crime and car break-ins are high
  • •Violent crime moderate
  • •Crime has declined since 2023
Overall:6.0
🏥

Healthcare

  • •Major hospitals: UCSF Health, Kaiser Permanente, Sutter CPMC
  • •World-class specialist care
  • •Access depends on employer health insurance; out-of-pocket costs can be high
Quality: 10.0Access: 7.0
🎭

Culture

  • •SFMOMA and de Young Museum
  • •LGBTQ+ history in the Castro
  • •Diverse food scene
9.0
🎉

Nightlife

  • •Mission District bars
  • •Live music venues
  • •Cocktail bars
8.0
🏞️

Recreation

  • •Golden Gate Park and Presidio
  • •Ocean Beach
  • •Marin Headlands hiking
9.0
💰

Cost of Living

  • •All amounts in USD: rent, utilities, groceries, internet, transit pass and gym per month; meals, coffee, beer and cinema per item
  • •Among the most expensive U.S. cities
  • •California income tax up to 13.3%
  • •Many apartments are rent-controlled
🏠 Rent (1BR center):€3400
⚡ Utilities:€256
🛒 Groceries:€570
🍽️ Dining out:€40

Engagement

Views
0today
7 days1
all-time1
Applies
0today
7 days0
all-time0
Black Forest Labs
Black Forest Labs
United States: San Francisco

Freiburg-based frontier AI research lab behind the FLUX family of image-generation and editing models (FLUX.1, FLUX.1 Kontext, FLUX.2 and successors), founded in 2024 by the researchers who created latent diffusion and Stable Diffusion. BFL ships both open-weight models (84M+ Hugging Face downloads) and a commercial API, licenses its models to large platforms, and is extending into multimodal models trained jointly on images, video, audio and action. Raised $300M Series B at a $3.25B valuation in December 2025.

EU Blue Card eligible
€per year
€20kthis job posts €130k – €240k€360k