Skip to main content
The Context Layer — a newsletter on AI memory, context synthesis, and building products that think.

Work

Every role has been a context architecture story — taking raw data, fragmented systems, or greenfield products and building the intelligence layer that makes them work.

Suzy
VP ProductDec 2025 – PresentDecision Engine
6 wks
concept → launch
350+
enterprise brands

The Challenge

Suzy had spent nine years as a market research platform trusted by 350+ enterprises — Microsoft, Google, PepsiCo, Netflix, P&G. But the product was a survey tool in a market being rapidly disrupted by AI. Enterprise marketing teams were drowning in fragmented intelligence across 20+ tools with no system that synthesized it all into action.

What I Did

Led product strategy for transformation to Decision Engine — structured around Intelligence, Insights, and Impact. Architected retrieval using Databricks vector DB with Qwen3-Embedding-8B. Designed AI evaluation framework. Drove inference cost optimization through model tiering (Opus/Sonnet/Haiku routing). Led GTM strategy. Reorganized product org from two pods into six triads shipping on daily cycles.

Context Connection

Suzy's three capabilities map directly to the five-step context generation framework: Intelligence is curation, Insights is synthesis and consolidation, Impact is prioritization and intelligent storage. The thesis showed up in the product before I named it.

Grandstage / Spade AI
Co-Founder & Head of Product2022 – 2024AI Market Intelligence
300%
user growth
$0
CAC
80%
retention

The Challenge

GTM teams at B2B companies were spending weeks compiling market insights from fragmented sources — analyst reports, news feeds, competitor data, social signals. The process was manual, expensive, and produced results that were stale by the time they reached a decision-maker.

What I Did

Led zero-to-one product development for a venture-backed platform ($525K raised). Managed ten across PM, UX, data science, and engineering. Shipped three AI products. Built hybrid search architecture: PostgreSQL + PGVector for dense retrieval, HNSW-indexed keyword search for sparse. Designed hierarchical relevance model using K-Means clustering. Custom evaluation and observability tooling increased precision over 40%.

Context Connection

This product worked because it consolidated raw market data into synthesized, goal-aware context — not because of the model. The architecture validated the five-step approach before I had the language for it.

Helm Labs
SVP & General Manager2024AI Data Intelligence
$3.25M
pipeline pre-launch
14x
partner ACV growth

The Challenge

Helm Labs (subsidiary of Murmuration, funded by Bloomberg LP) had acquired five separate products serving political and public affairs professionals. They needed a unified platform strategy, a coherent product roadmap, and a go-to-market motion built from scratch — for a product that didn't exist yet.

What I Did

Defined enterprise data platform vision integrating proprietary datasets and custom ML models covering 200+ behavioral and demographic data points on ~200M Americans. Built GTM strategy selling platform access with research services. Led search interface redesign reducing response time 50%+. Achieved 105% gross margin growth through pricing optimization.

Context Connection

The value I sold was context architecture — how information about ~200 million Americans would be prioritized and presented to decision-makers. The pipeline was built on the strength of the vision and the services layer, not a shipping product.

IBM
Digital Product & Growth Leader2018 – 2021Growth — Cloud & Watson
31%
trial conversion lift
$2.4M
e-nurture revenue
8 fig
pipeline

The Challenge

IBM's Cloud and AI business needed to fundamentally transform how it acquired, converted, and retained approximately one million trial users globally. The digital experience was generic — same content and journey for every user regardless of product interest, stage, or behavior.

What I Did

Oversaw 48 reports and $4M budget. Built content authoring features for IBM's DXP cutting time-to-market 50%+ and doubling organic traffic. Launched IBM's first personalization pilot. Developed predictive dashboards and automated personalized onboarding. Deployed growth stack (Braze, Segment, Amplitude) across 30+ countries with 100% adoption.

Context Connection

Personalization at scale is a context problem: what does this specific customer need to see, right now, to make a decision? The systems I built at IBM were early context-aware architectures applied to growth.

Phase2
Director, Product & Digital Strategy2014 – 2017Enterprise Transformation
$3M
practice revenue yr 1
500%
J&J revenue growth

The Throughline

Launched and led a 10-person digital and product strategy consulting practice. Embedded as product lead within major enterprise clients, each time transforming how the organization used digital platforms and intelligent tools.

Johnson & Johnson

Overhauled CMS powering 1,000+ product websites. Cut time-to-market from months to under four weeks. Grew engagement revenue 500%+.

Northwell Health

Designed editorial platform transforming patient engagement. Led CMS enhancements and decoupled React frontend. Reactivated tens of thousands of dormant patients.

Ann Inc. (Ann Taylor)

Managed design and build of digital workspace for retail employees across brands.

Other Platforms

Twitter international SEO tools. Memorial Sloan Kettering CMS. Reddit digital support center. Al Jazeera omnichannel CMS. Sony Music collaboration system.

Meet the person behind the work

About me