Head of AI · LLM Platform & Decision Systems Leadership
PHILIP CHEUNG
AI-first technology leader who ships production LLM systems end-to-end: strategy, architecture, governance, and runtime.
Head of AI at Weidmann & Cie. AG. Leads Qorinix, LaSpend, and Fixxmi platform delivery with multi-provider AI routing and audit-grade decision workflows.
MSc Computer Science, AI & Data Science (Merit)
- Builds enterprise AI platforms that connect business goals to deployable systems
- LLM routing · RAG · guardrails · human-review gates · sub-100ms inference engineering
- Leads cross-functional AI teams across product, compliance, and engineering
Professional Summary
Philip Cheung is an AI-first technology leader focused on turning LLM capability into measurable business outcomes. He operates at the Head of AI level across strategy, architecture, execution, and governance.
He builds production AI systems for high-stakes environments: multi-provider model routing, policy-aware prompt orchestration, RAG with evidence traceability, human-review escalation, and cost/performance telemetry.
At Weidmann & Cie. AG, he leads Qorinix, LaSpend, and Fixxmi AI platform execution, combining low-latency inference engineering with auditability and compliance-ready controls. He regularly aligns engineering design with legal, risk, and operations stakeholders.
His background spans AI engineering, platform design, and technical leadership across cloud-native systems. He is strongest when a company needs both AI depth and delivery discipline to move from prototype to resilient production.
Core Competencies
Head of AI Leadership
AI roadmap, org design, operating model, cross-functional execution
LLM Platform Architecture
Model routing, prompt registry, RAG pipelines, reliability engineering
AI Governance & Safety
Guardrails, policy enforcement, audit trail, human-in-the-loop approvals
ML/Inference Engineering
Sub-100ms serving, scaling, observability, continuous evaluation
Product & Delivery
Translate business priorities into shipped AI capabilities and KPIs
Stakeholder Communication
Executive narrative, hiring support, architecture reviews, mentoring
Notable Achievements
Head of AI Leadership
Defined and executed AI roadmap across platform, governance, and delivery. Built an operating model where engineering, product, compliance, and business stakeholders align on one decision system.
Qorinix LLM Program
Led end-to-end design, training, and deployment of Qorinix LLM for production decision support. Focused on reliability, latency, and auditability in a high-stakes operating context.
Multi-Provider AI Platform
Delivered model routing, fallback policy, prompt registry, and evaluation workflow so teams can adopt AI safely without locking into a single provider.
Audit-Grade AI Governance
Implemented policy checks, evidence traceability, and human-review escalation for regulated decision flows. Every material decision can be traced, reviewed, and explained.
Inference & Runtime Performance
Architected low-latency inference services with observability and failover patterns, enabling stable production behavior under live workload pressure.
Team Building & Mentorship
Mentored engineers and set technical standards for AI delivery, including model lifecycle management, release discipline, and cross-functional review rituals.
Research Foundation
MSc in AI & Data Science (Merit), with work spanning LLM systems, machine learning, and production-grade AI architecture.
Business Impact
Converted AI from experimentation to operational capability with measurable impact on decision speed, engineering quality, and organisational confidence in AI rollout.
Professional Experience
Weidmann & Cie. AG
London, UK
Responsibilities
- Owns the AI strategy and execution roadmap across Qorinix, LaSpend, and Fixxmi product lines.
- Built a multi-provider LLM platform with routing, fallback policy, prompt registry, and cost/latency observability.
- Established production RAG workflows with evidence packaging and policy checks for high-stakes decisions.
- Implemented guardrails and human-review escalation patterns for regulated or sensitive workflows.
- Led architecture reviews and translated leadership priorities into delivery plans and release gates.
- Shipped low-latency inference patterns with reliability controls, canary rollout, and rollback playbooks.
Key Outcomes
- Operationalised AI decision workflows with audit traceability and governance checkpoints.
- Standardised model lifecycle practices: evaluation baselines, release criteria, and drift monitoring.
- Improved cross-team AI delivery speed by introducing reusable platform capabilities instead of one-off integrations.
Skills & Technologies
Pacific Cloud Computing Ltd.
Hong Kong & Remote UK
Responsibilities
- Led AI transformation and built intelligent systems processing millions of predictions daily.
- Designed distributed inference and data pipelines with production reliability targets.
- Built RAG and NLP systems for faster information retrieval and richer decision support.
- Introduced MLOps practices: retraining workflow, staged rollout, and continuous quality checks.
- Mentored engineers and established architecture and coding standards.
Key Outcomes
- Scaled ML services into stable production operations.
- Reduced retrieval latency and improved information quality for internal teams.
- Delivered enterprise systems with high availability and operational discipline.
Skills & Technologies
Groupon.com
Hong Kong
Responsibilities
- Led strategic account management and merchant coordination.
- Managed project planning, estimation, and execution tracking.
- Improved operational process and delivery quality.
Key Outcomes
- Delivered measurable revenue and operational improvements.
Skills & Technologies
Flagship AI Programs
A curated view of AI leadership, LLM platform architecture, and production decision systems delivered by Philip.
Head of AI Operating Model
Defined an execution model that aligns product, engineering, compliance, and business stakeholders around one AI roadmap. Built delivery rhythms, review gates, and accountability paths so AI work moves from prototype to reliable production.
Qorinix LLM, In-House Program Leadership
Led end-to-end design, training, and deployment of Qorinix LLM. Emphasis was on controlled runtime behavior, policy-safe decision support, and measurable latency/throughput gains for high-stakes workflows.
Multi-Provider LLM Routing Control Plane
Built model routing and fallback policy across providers, with prompt versioning and performance/cost visibility. This reduced lock-in risk and allowed teams to scale usage without fragile one-off integrations.
Audit-Grade AI Decision Loop
Designed closed-loop decision flow with policy checks, evidence retrieval, action runtime, and audit ledger. Every critical decision is traceable, reviewable, and aligned with governance requirements.
Sub-100ms Inference Mesh
Engineered low-latency inference services with routing, autoscaling, and health-weighted failover. Supports high-volume production usage while maintaining stable performance and operational resilience.
Real-Time AI Operations Dashboard
Built operational observability across model performance, cost, policy events, and service reliability. The dashboard enables faster incident response and executive-level reporting on AI system health.
Unified AI Delivery Stack
Integrated Python, TypeScript, FastAPI, Next.js, PostgreSQL/TimescaleDB, Redis, Docker, and cloud services into a cohesive AI delivery stack. Designed to support both rapid iteration and production-grade controls.
LLM-Augmented Research to Production Pipeline
Converted advanced research concepts into deployable systems through evaluation discipline, release gates, and runtime monitoring. Focus on repeatability and business impact, not one-off experiments.
Skills & Expertise
Programming & Frameworks
AI/ML Capabilities
AI Leadership & Strategy
LLM Platform & Runtime
Backend & Data
Cloud & Delivery Tooling
AI System Design
Domain & Stakeholder Skills
Education & Certifications
Education
AI Systems, Applied ML, and LLM Decision Workflows
University of Wolverhampton
Visit WebsiteFocused on production-oriented AI, machine learning system design, and evidence-based decision support architecture.
Marketing (with IS minor)
The Hong Kong University of Science & Technology
Visit WebsiteFoundation in business strategy, communication, and cross-functional leadership.
Certifications & Qualifications
Applied AI enablement and practical adoption coaching.
Google Cloud
Visit WebsiteProduction lifecycle practices for ML systems.
Harvard University
Visit WebsiteComputer science foundations and software engineering discipline.
SFC / HKSI
Visit WebsitePassed Papers 1, 7, 8, 12.
Frequently Asked Questions
Common hiring and execution questions about Philip's Head of AI profile
Resources & Playbooks
A practical framework for first-90-day AI leadership: baseline assessment, platform foundation, and high-value use-case delivery plan.
Policy-first architecture for LLM deployments, covering prompt controls, human-review escalation, and auditability requirements.
How to set measurable release criteria, benchmark drift, and maintain quality over time in production AI systems.
Loading booking calendar...
If the embedded widget is blank, please use direct booking.
Open CalendlyContact Me
Use contact form only
Phone
Use contact form only
Location
London, United Kingdom