Developer Marketer · AI observability
Tori 1.0 — hire me.
I didn't send a résumé into the void. I shipped a real AI feature into my own production pipeline, instrumented it in PostHog AI observability, and used evals to find exactly where it broke. This page is the cover letter. It's also tracking you in PostHog right now — scroll down, you'll see it. (Transparent by default. Felt on-brand.)
I didn't apply. I shipped.
The role is Developer Marketer for AI observability. So instead of describing the product, I used it. I built card-finish detection for my Pokémon TCG pipeline (ChaseDex) — a vision model that reads whether a card is normal, holo, or reverse-holo, which is what sets its price — and wired every model call through PostHog AI observability. Then I evaluated it against my own labeled scans.
The eval loop, run live
Three runs. Each one a trace in PostHog. The numbers aren't the point — the diagnosis is.
| Run | Change | Accuracy | What the data revealed |
|---|---|---|---|
| v1 | naive prompt | 42% | Model called everything reverse — it equated scanner glare with foil. |
| v2 | "glare ≠ foil" prompt | 47% | normal recall doubled. The fix landed exactly where the bias was. |
| v3 | full resolution | 53% | holo still 0% — even at full res. |
The finding: holo is undetectable from flat scans regardless of prompt or resolution — foil only reveals its pattern under angled light. Evals didn't just tune a prompt; they told me the limiting factor is data capture. That's the entire value of AI observability, in one true story.
And here's how I'd sell it
Every competitor can only see the AI. PostHog sees the AI in the context of the product and the user — because analytics, session replay, flags, and experiments are already in the same project. You can go from "this generation was slow, expensive, or wrong" to "…and here's the user it happened to, what they did next, and whether they churned." Nobody else can close that loop. Full teardown — positioning vs Datadog / LangSmith / Langfuse, the launch plan, and a GEO program for winning "what should I use to monitor my LLM app?" in ChatGPT and Perplexity.
Why me, quickly
- 6 years, all developer tools. PLG + technical GTM, nothing else.
- Depot: $3M → $11.5M ARR. Built the growth engine from scratch; leads 139 → 707.
- AI search visibility is my edge. GEO/AEO across ChatGPT, Claude, Perplexity, Kagi.
- I ship code. This demo, an iOS app, internal tooling — JS/Python, APIs, CLIs.
- I live in PostHog. Built attribution + cohorts in it at Depot; the traces behind this page are in my own project.
Things I've shipped
Marketing leader who codes. I build and run real products — product, engineering, and growth, end to end. A sampling:
ChaseDex
Pokémon TCG collection tracker
An iOS app (live on the App Store), an SEO/affiliate website, and a daily data pipeline that powers both.
React Native/Expo · Cloudflare Workers + R2 · PostHog
DropSync
Pokémon drop-timing tool
A live product-drop calendar plus a millisecond-accurate refresh "sniper" Chrome extension — with accounts and Stripe billing.
Chrome MV3 extension · Cloudflare Workers / KV / D1 · Stripe
SmartHair
AI hair-care iOS app
Personalized AI recommendations with subscription billing — shipped solo, product through growth.
React Native/Expo · RevenueCat · custom AI
The résumé
Download PDF ↓- Joined at $3M ARR; built the growth engine that drove the business to $11.5M ARR (path to $30M).
- Monthly leads 139 → 707 (5.1×), paid acquisition 34 → 137 (4×) in year one, across a $1.3M budget.
- Led GTM launch of Depot CI; built a developer PLG engine (technical content, creator programs, AI-search visibility, lifecycle).
- Built attribution + cohort analysis systems in PostHog — the company's first clear view of what drove revenue.
- Generated $1M inbound pipeline in a single month via technical product-positioning campaigns.
- Improved marketing efficiency 85%; organic conversion 16.75% → 27.65% via technical SEO + CRO.
- Built full-funnel attribution connecting product usage, acquisition, and sales pipeline.
- Email campaigns at 65% open rates across 20,000+ attendees; partnered with C-level sponsors (G2, ZoomInfo, Vidyard).
- Improved lead retention 65%, consistently beat quota; authored reporting + campaign playbooks.
PLG acquisition · developer marketing · technical content · lifecycle & activation · demand gen · creator/influencer programs · AI search visibility (GEO/AEO) · attribution modeling · cohort analysis · PostHog · HubSpot · A/B testing · CRO · JavaScript · Python · GitHub/APIs/CLI · shipped iOS app · $1.3M budget ownership
B.S., Communication Sciences and Disorders — Illinois State University. (Selected projects above.)