STEM
FilterForge
Design a chat-safety system and measure it honestly (ages 15–18): on a fully synthetic corpus, set the rules and threshold that flag grooming, off-platform lures, and age-band mismatch — then confront the two error costs (a false alarm silences a child talking to a cousin; a miss lets the lure through) on your own precision/recall curve before the engine reveals the truth. Trust-and-safety engineering. Deterministic, on-device; text-forward.
#0e7490 Distributed-narrative cast
Meet the cast
FilterForge's adapted-DN-S cast (ages 15–18, realistic trust-and-safety personas — no mascots, per R-OLDER-TEEN-DN-ADAPTED) each embody one idea in honest measurement; mentor Vale asks what a single accuracy number hides.
Vale
(Mentor) — asks what one accuracy number hides about the two error costs
Prei
precision — of everything you flagged, how much was truly harmful
Reccia
recall — of all the real harm, how much you actually caught
Thresh
the threshold — the dial that trades a miss against a false alarm
Cousin Cora
the false-alarm cost — a child talking to a real cousin, wrongly silenced
What's inside
Learning goal
Design a chat-safety system and measure it honestly (ages 15–18): on a fully synthetic corpus, set the rules and threshold that flag grooming, off-platform lures, and age-band mismatch — then confront the two error costs (a false alarm silences a child talking to a cousin; a miss lets the lure through) on your own precision/recall curve before the engine reveals the truth. Trust-and-safety engineering. Deterministic, on-device; text-forward.
Question kits
16 curriculum-aligned kits × 25 questions = 400 questions per app, mapped to recognized standards.
On-device AI mentor
FoundationModels-powered hints, feedback, and adaptive difficulty — all running locally.
Mentored by Vale — on-device AI, no data leaves the device.
How FilterForge handles your kid's data
- ✅ All progress, settings, and AI-generated content stays on the device
- ✅ No analytics, no tracking, no third-party SDKs
- ✅ No ads, no in-app purchases — you pay once
- ✅ COPPA compliant under the 2026 FTC amendments
- ✅ Parental controls + session limits + content filters built in
FilterForge runs on ForgeKit — the open-source Swift Package Manager framework that powers every Spark & Anvil app. ForgeKit ensures consistent accessibility, COPPA compliance, and design language across the portfolio, so your kid's progress and preferences feel coherent across every app they touch.
Coming to the App Store
FilterForge is in active development. Email us to hear when it ships — no marketing, no spam, just a one-shot launch announcement.
Email me at launch