About
Max co-founded Lyceum to build intelligent GPU orchestration software. He holds a PhD in Theoretical Machine Learning from the University of Cambridge.
Before Lyceum, Max was a Senior Data Scientist at QuantumBlack (McKinsey's AI division), Head of Machine Learning at a hyperspectral imaging startup, and a Quant Trader at Citadel. His research background spans computational biology, chaos theory, and deep learning.
At Lyceum, Max leads the technical vision - building the AI systems that match workloads to hardware automatically.
Published Articles
LLM Inference & Model Serving
- • Deploy Gemma 3 on European GPU Cloud: VRAM, Setup, and GDPR Compliance
- • vLLM Production Deployment Guide: Scaling Sovereign Inference
- • The Economics of Scale to Zero: Slashing GPU Inference Costs in 2026
- • NVIDIA Dynamo 1.0: A Technical Guide to Inference Orchestration
- • Deploying Private LLM Endpoints on GPU Cloud: A 2026 Strategy
- • Dedicated vs Shared GPU Inference: Scaling AI Infrastructure
Production GPU Infrastructure
GPU Cost Optimization
- • Multi-Cloud GPU Strategy: How to Avoid AI Infrastructure Vendor Lock-In
- • GPU Idle Cost Waste Calculator: Stop Paying for 5% Utilization
- • H100 vs B200 GPU Cost Efficiency Comparison for AI Workloads
- • NVIDIA B200 GPU Cloud Pricing 2026: True Costs & Architecture
- • NVIDIA B200 vs H200 GPU for Inference: Architecture & Benchmarks
Sovereign AI Infrastructure
GPU Infrastructure & Cost Engineering
GPU Cloud Migration & Alternatives
EU-Sovereign AI Compute
GPU Memory Management
- • NVIDIA B200 192GB VRAM Model Requirements: A Technical Guide
- • ZeRO-3 vs FSDP: A Deep Dive into Memory Efficiency for LLMs
- • KV Cache Memory Calculation for LLMs: A Technical Guide
- • How Much VRAM for a 70B Model? A Technical Engineering Guide
- • Maximizing VRAM: Gradient Checkpointing Memory Savings Guide
- • PyTorch Memory Profiling in Production: A Guide to Efficiency
- • How to Predict VRAM Usage for PyTorch Models
- • Solving OOM Errors in 70B Model Fine-Tuning
- • How to Prevent OOM Errors in PyTorch Training
- • GPU Utilization Too Low: How to Fix Compute Bottlenecks
- • GPU Memory Estimation: A Guide to VRAM Requirements
- • GPU Memory Calculator for Deep Learning: A Technical Guide
- • Solving CUDA Out of Memory Errors in Llama Fine-Tuning
- • Eliminating CUDA OOM: Expert Memory Management for LLMs
Want to join the team?
We're always looking for exceptional engineers and builders who care about infrastructure.