Strategic AI Advisory

AI systems that work for organizations that matter

Helping mission-driven companies build the right AI — from strategy through production. Founded by a principal AI scientist with a decade building systems for humanitarian response, global development, and enterprise scale.

Start a conversation

About

From the field to production AI — and back again

I started my career doing fieldwork in Southeast Asia and West Africa — biodiversity surveys, value chain assessments, the kind of work where you're in a truck at dawn and presenting to a ministry by afternoon. That led to a PhD, collaborations with NASA and the UN FAO, and a chapter in the World Bank's flagship report on data for policy.

Then I pivoted to industry. Over the past five years I've built production ML and AI systems at scale — agentic AI processing 500K+ daily calls, distributed video pipelines on Kubernetes, RAG systems, voice AI, LLM evaluation infrastructure serving 100M+ users. I've led teams, cut millions in costs, and shipped systems that work.

Now I'm bringing those capabilities back to where they matter most. AI isn't something to fear or pause — it's here, and the organizations working on health, climate, education, and humanitarian response need senior technical guidance to use it well. That's what Glide AI does.

10+
Years in AI & ML
40+
Publications
2,200+
Citations
155
Countries impacted

Services

What Glide AI delivers

01

AI Strategy & Architecture

Which models should you use? Build or buy? What does an agentic AI pipeline actually look like for your use case? I help founders and technical leaders design AI systems that are right-sized for their stage, their data, and their mission — from LLM selection and RAG architecture to voice AI and evaluation frameworks.

02

Technical Hiring & Team Building

Writing JDs that attract the right candidates. Interviewing data science and AI leadership. Building hiring roadmaps so you know who to bring on first, second, third — and what to outsource. Especially valuable for early-stage teams that don't yet have senior AI leadership in-house.

03

Project Scoping & Roadmapping

Turning a vague "we want to use AI" into a concrete, phased plan with realistic timelines and resource requirements. I help you figure out what's feasible, what's not, and where to start — so you don't burn six months and a seed round learning the hard way.

Selected Work

Track record

Asurion — Enterprise

Agentic AI Call Intelligence

Led production agentic AI systems processing 500K+ daily calls with real-time voice AI, LLM evaluation, and adversarial safety testing at scale.

$2.2M annual savings

Canvs AI — Enterprise

LLM Pipeline at Scale

Built production RAG-LLM pipeline processing 1.5M+ customer feedback texts daily. Improved system accuracy from 45% to 90% through evaluation infrastructure.

90% accuracy, 80% cost reduction

Premise Data — Humanitarian

Crisis Intelligence Platform

Architected ML systems for real-time market intelligence across 155 countries — distributed video pipeline processing 100K+ videos/day.

$3M in revenue & savings

World Bank — Global Development

Data for Better Lives

Co-authored chapter in the World Development Report on big data for policy. Lead data scientist for the SDG Atlas.

~1M downloads

NASA & UN FAO — Research

Global-Scale ML Models

Built predictive models from satellite and sensor data at 10km global resolution. Food security monitoring with uncertainty quantification.

2M+ time series

Gates Foundation — NLP

Research Synthesis at Scale

Fine-tuned BERT to synthesize 18K research articles for evidence-based agricultural policy, achieving 88% classification accuracy.

18K articles processed

AI isn't something to pause or fear — it's here. The question is whether the organizations doing the most important work in the world have access to senior technical guidance. That's the gap Glide AI fills.

How It Works

Strategic advisory, not outsourced engineering

I work as an embedded strategic advisor — the senior AI perspective your team needs without a full-time executive hire. Engagements are scoped tightly so you get clarity and direction without scope creep.

This isn't implementation work. I'm not writing your production code. I'm telling you what to build, how to architect it, who to hire to build it, and what pitfalls to avoid — based on having built these systems myself at enterprise scale.

Typical engagement

Most advisory relationships run at roughly 10 hours per month — enough for regular working sessions, architecture reviews, hiring support, and async guidance. Engagements are structured around clear deliverables with defined timelines, not open-ended retainers. If you have a specific project that needs scoping, we can structure a shorter engagement around that.