2021
Trillbit Inc
trillbit.com
Background
In March 2020, At Trillbit, we were watching every contact-tracing app hit the same wall: they only warned users after exposure. A pre-print from Carnegie Mellon on “network-distance” alerts showed a different path: predict risk before contact.
Motivated by a desire to empower individuals with timely information, we conceptualised Notch—a novel approach to contact tracing that aimed to alert users before exposure, leveraging network theory and advanced proximity detection technologies.
Core problem
Traditional contact tracing methods, both manual and digital, primarily focused on notifying individuals after they had been in close contact with someone who tested positive for COVID-19. This reactive approach often resulted in delayed interventions, allowing the virus to spread unchecked.
Pain Point | Why it mattered |
|---|---|
Lagging alerts: BLE / Wi-Fi apps notified users days after exposure | Time lost meant new infection chains. |
Unreliable distance: RSSI swings 1–2 m with pockets, walls and phone models | False positives drained trust and compliance. |
Hidden CAPEX: Reaching sub-meter BLE accuracy needs dense beacons and extra Wi-Fi APs | During a pandemic downturn most venues could not fund new infrastructure. |
Privacy fears: App sign-ups asked for emails or GPS, depressing adoption | Low network effects crippled usefulness. |
Solution
Notch fused three complementary signals already present in every modern smartphone, then layered a graph-distance model (“You are 2 hops from a positive case”) on top.
Layer | Why we included it | How we use it |
|---|---|---|
Ultrasound | Sub-meter precision in quiet environments | Primary distance measure within ~3 m; phone chirps are inaudible to users. |
BLE | Works through cloth / pockets where ultrasound may fail | Fills gaps; used to corroborate or flag suspect readings. |
Wi-Fi RTT / RSSI | Longer-range context indoors (5-10 m) | Extends the network graph so we can warn users multiple “hops” away even when they never exchanged chirps. |
Key advantages
No new hardware: We leverage the radios you already own—no beacons, no extra APs.
Confidence scoring: Sensor fusion algorithm weights each signal by environment (library vs. factory floor) to cut false alerts by 42 % in pilot tests.
Same privacy promise: Phones exchange rotating, anonymous IDs; no GPS, no PII, ever.
My Role
Product Strategy — Led user research and defined the core "pre-exposure" hop-based model, translating insights into PRD, JTBD mapping, and OKRs.
Design — Sole product designer; built a clean, minimalist UI system in Figma, designing all flows across web, iOS, and Android for onboarding, risk visualization, and settings.
Technical Architecture — Spearheaded the BLE + Ultrasound + Wi-Fi sensor fusion design; built early proximity modules using BLE and Trillbit SDK
Execution & Team Leadership — Shipped TestFlight MVP in 6 weeks
Privacy & Compliance — Designed privacy-first architecture with anonymous rotating IDs; passed GDPR and Indian DPDP audits with zero findings.
Results and Impact
Notch was built to rethink contact tracing from the ground up.
Powered by a fusion of ultrasound, Bluetooth, and Wi-Fi, Notch achieved sub-meter accuracy without requiring any external hardware like BLE beacons or additional Wi-Fi APs.
Our goal: deliver pre-exposure alerts that were early, precise, and privacy-first — fixing the failures of Bluetooth-only or Wi-Fi-only systems.
Results & Key Metrics — Notch Deployment at GetLook
Social Distancing Adherence
98% adherence achieved, compared to 45% when manually enforced by the company.
Customer Job Assignment Decisions
100,000+ customer job assignments made based on daily risk reports generated by Notch.
Risk Profiling and Quarantine Decisions
300+ risk profiling-based decisions made for quarantine recommendations, helping proactively manage workforce exposure.
Business Continuity Impact
Enabled GetLook (India’s leading home salon service) to serve 100,000+ customers safely during peak COVID months without compromising employee or customer health.
Timely Decisions
Over 30+ critical quarantine or work assignment decisions made in real-time using Notch’s risk profiling insights.
