AI-powered E-commerce for Eyewear with Virtual Try-On
Challenge
Specific challenges defined the project scope:
Develop a virtual try-on feature using AI and facial recognition
Customers needed to see how frames looked on their face before purchase. The technology had to work reliably across different device cameras, lighting conditions, and facial geometries.
Improve website navigation and interface for product discovery
Lenskart's catalog covers thousands of frames across prescription eyewear, sunglasses, contact lenses, and accessories. Customers were struggling to navigate, filter, and compare effectively.
Integrate popular payment methods for fast, secure checkout
Online checkout abandonment is a known issue in e-commerce. Lenskart needed a payment layer supporting multiple gateways and methods to reduce friction at the final step.
Ensure a smooth product browsing experience with abundant eyewear options
The site had to handle large catalog volumes without sacrificing page load speed or browse fluidity, especially on mobile where most traffic arrives.
Specific challenges defined the project scope:
Develop a virtual try-on feature using AI and facial recognition
Customers needed to see how frames looked on their face before purchase. The technology had to work reliably across different device cameras, lighting conditions, and facial geometries.
Improve website navigation and interface for product discovery
Lenskart's catalog covers thousands of frames across prescription eyewear, sunglasses, contact lenses, and accessories. Customers were struggling to navigate, filter, and compare effectively.
Integrate popular payment methods for fast, secure checkout
Online checkout abandonment is a known issue in e-commerce. Lenskart needed a payment layer supporting multiple gateways and methods to reduce friction at the final step.
Ensure a smooth product browsing experience with abundant eyewear options
The site had to handle large catalog volumes without sacrificing page load speed or browse fluidity, especially on mobile where most traffic arrives.
Solution
Research-led development with AI capability as a first-class feature.
1. Discovery and technology selection
Before any development started, we evaluated facial recognition libraries and computer vision approaches against the constraints of running in a customer's browser. Most virtual try-on solutions on the market either required app installation (high friction) or compromised on accuracy. We selected the technical approach that balanced accuracy with web-first delivery — no app download required, working across major browsers and device cameras.
2. UX redesign anchored on product discovery
Improving navigation and product filtering is straightforward as a feature list. It becomes harder when the catalog has thousands of frames and the customer does not know what they are looking for. We restructured filtering around the questions customers actually ask (face shape, frame style, prescription type, brand, price) rather than the catalog's internal data model. The navigation redesign followed the same principle — organized by intent, not by inventory category.
3. Payment layer with conversion focus
Payment integration in e-commerce is often treated as a checkbox. We treated it as a conversion problem. The implementation handles multiple payment methods (PayPal, Visa, Mastercard, regional alternatives), retries failed transactions, and minimizes the number of steps between adding to cart and completing payment. Each payment integration was added with proper error handling and reconciliation reporting — the unglamorous parts of payment systems that determine whether the system is usable in production.
Research-led development with AI capability as a first-class feature.
1. Discovery and technology selection
Before any development started, we evaluated facial recognition libraries and computer vision approaches against the constraints of running in a customer's browser. Most virtual try-on solutions on the market either required app installation (high friction) or compromised on accuracy. We selected the technical approach that balanced accuracy with web-first delivery — no app download required, working across major browsers and device cameras.
2. UX redesign anchored on product discovery
Improving navigation and product filtering is straightforward as a feature list. It becomes harder when the catalog has thousands of frames and the customer does not know what they are looking for. We restructured filtering around the questions customers actually ask (face shape, frame style, prescription type, brand, price) rather than the catalog's internal data model. The navigation redesign followed the same principle — organized by intent, not by inventory category.
3. Payment layer with conversion focus
Payment integration in e-commerce is often treated as a checkbox. We treated it as a conversion problem. The implementation handles multiple payment methods (PayPal, Visa, Mastercard, regional alternatives), retries failed transactions, and minimizes the number of steps between adding to cart and completing payment. Each payment integration was added with proper error handling and reconciliation reporting — the unglamorous parts of payment systems that determine whether the system is usable in production.
Results
Eyewear is the rare retail category where customers physically need to see how a product looks on them before committing. The virtual try-on feature removed this barrier in Lenskart's online flow for the first time, letting customers complete the journey online without compromising on the confidence of trying frames in person. Combined with restructured filtering that matches how customers actually shop and a checkout flow designed to convert, the platform turned a large catalog into a manageable browsing experience and reduced friction at the final purchase step.
The technology layer Adamo helped build operates at significant scale and continues to evolve as Lenskart expands. The company now runs more than 2,000 retail stores across India and has moved internationally into Singapore, Japan, Thailand, and the Middle East, backed by investors including SoftBank, KKR, and Temasek. Adamo's work on virtual try-on, product discovery, and checkout flexibility was one part of a much larger Lenskart technology investment, focused on the customer-facing layer where the in-store experience needed to translate online.
Eyewear is the rare retail category where customers physically need to see how a product looks on them before committing. The virtual try-on feature removed this barrier in Lenskart's online flow for the first time, letting customers complete the journey online without compromising on the confidence of trying frames in person. Combined with restructured filtering that matches how customers actually shop and a checkout flow designed to convert, the platform turned a large catalog into a manageable browsing experience and reduced friction at the final purchase step.
The technology layer Adamo helped build operates at significant scale and continues to evolve as Lenskart expands. The company now runs more than 2,000 retail stores across India and has moved internationally into Singapore, Japan, Thailand, and the Middle East, backed by investors including SoftBank, KKR, and Temasek. Adamo's work on virtual try-on, product discovery, and checkout flexibility was one part of a much larger Lenskart technology investment, focused on the customer-facing layer where the in-store experience needed to translate online.