What I Offer

I partner with product and technical teams to navigate the early, high-impact stages of AI development—where direction matters most.

Where I Add the Most Value

• 🧠 Solving Hard Problems: Tackling complex AI challenges that go beyond standard tools.
• 🧬 Cross-Disciplinary Innovation: Combining ideas from language models, reinforcement learning, biosignals, and more.
• ⚙️ Feasibility & Prototyping: Quickly validating concepts, designing systems, and building key prototypes.
• 🧑‍🏫 Technical Coaching & Upskilling: Helping teams sharpen their design, architecture, and problem-solving skills.

What I Don’t Do

• ❌ Long-term implementation or ongoing maintenance
• ❌ Commodity AI development work
• ❌ Full-time engineering or MLOps roles

Areas of expertise

📘 1. LLM Strategy & Architecture

Designing practical, well-scoped AI features

• Clarifying use cases and assessing feasibility
• Choosing between custom or API models
• Designing architectures: prompting, retrieval, orchestration
• Defining success metrics and evaluation methods
• Prototyping key components or guiding your team through the process

🔍 2. Search & Recommendation Systems

Building smarter search and personalized recommendations that drive real results

• Semantic search and hybrid ranking
• Personalized recommendations
• User modeling, scoring, relevance metrics
• Prototyping ranking models and pipelines
• Auditing and improving existing systems

🌐 3. Cross-Domain AI Innovation

Exploring bold AI ideas by combining techniques from different fields

• Reinforcement learning for optimization and decision-making
• Deep learning on biosignals and complex data
• Custom metrics and system evaluation
• Fusion of AI methods across domains

Example Engagements

⚡ 1. AI Innovation Sprint

4-7 weeks to go from early-stage idea to a working prototype — fast, but thoughtful.

Phases:

• Discovery & Ideation – Spot high-impact opportunities
• Design & Feasibility – Choose architecture, models, and metrics
• Prototyping – Build key components and support your team

🔍 2. Technical Audit & Solution Blueprint

2–4 weeks to uncover blockers and deliver a clear, actionable path forward.

Phases:

• 🧪 Audit – Review code, models, metrics, and workflows
• 🚧 Root Cause Analysis – Pinpoint issues and risks
• 🗺️ Solution Blueprint – Provide a roadmap and Go/No-Go decision

🧑‍💼 3. Fractional AI Architect

Ongoing support to guide strategy, design, and critical technical decisions — without a full-time hire.

• Architecture and design reviews
• Core algorithm and model coaching
• Hands-on prototyping support
• Metrics design and interpretation
• Monthly alignment with leadership