How can we enhance Onfido's facial recognition technology to improve accuracy in low-light conditions?

Introduction

To enhance Onfido's facial recognition technology for improved accuracy in low-light conditions, we need to approach this challenge systematically. This improvement is crucial for expanding the technology's applicability and reliability across various environments. I'll outline my approach to tackle this product improvement, focusing on user needs, technical solutions, and measurable outcomes.

Step 1

Clarifying Questions (5 mins)

• Looking at the product context, I'm thinking about the primary use cases for Onfido's facial recognition in low-light conditions. Could you provide more insight into the specific scenarios where users are encountering this issue most frequently?

Why it matters: Helps prioritize which low-light environments to optimize for first. Expected answer: Late-night identity verification for ride-sharing or food delivery services. Impact on approach: Would focus on mobile-optimized solutions for outdoor, urban environments.

• Considering user behavior, I'm curious about the current user experience when attempting facial recognition in low light. What's the typical failure rate or user frustration point in these conditions?

Why it matters: Determines the severity of the problem and sets a baseline for improvement. Expected answer: 30% failure rate in low light, with users often needing 3+ attempts. Impact on approach: Would prioritize solutions that can dramatically reduce first-attempt failures.

• Examining the product lifecycle, where does facial recognition accuracy sit in terms of Onfido's overall product strategy? Is this a core differentiator or a necessary feature to keep pace with competitors?

Why it matters: Helps align the solution with broader company objectives. Expected answer: Core differentiator, critical for maintaining market leadership. Impact on approach: Would justify more significant investment in cutting-edge technology.

• Considering external factors, how has the increasing prevalence of remote work and 24/7 digital services affected the demand for accurate low-light facial recognition?

Why it matters: Provides context for the urgency and potential market impact of the improvement. Expected answer: Significant increase in off-hours verification needs across industries. Impact on approach: Would emphasize scalability and adaptability to diverse lighting conditions.