Constructor
What is Constructor?
Constructor is an ecommerce discovery platform for merchandising and digital commerce teams that personalizes search, browse, recommendations, and shopping agents across the storefront. Its Commerce Reasoning Engine uses behavioral clickstream and shopper data with reinforcement learning to improve in real time. Key features include AI Shopping Agent, Product Insights Agent, Search & Autosuggest, and Merchandiser Controls. It powers brands such as Sephora, Under Armour, and REI.
Last verifiedHow we evaluate
At a glance
- Constructor is best for ecommerce teams that need personalized discovery across search, browse, and shopping agents.
What does Constructor do?
Constructor's Commerce Reasoning Engine turns behavioral clickstream and petabytes of shopper and market data into personalized discovery across the storefront. It uses reinforcement learning to keep improving in real time, so search, browse, recommendations, and shopping agents can adapt as shoppers interact. The result is a connected discovery system rather than isolated tools: AI Shopping Agent guides open-ended queries, Product Insights Agent answers product questions, and Search & Browse shapes results and category pages around intent. At scale, Constructor says it serves billions of queries every day and powers discovery for brands such as Sephora, Under Armour, Gap, REI, and Tractor Supply. The platform is built with security and privacy in mind, with minimal personal data collection, SOC 2 Type 2 and ISO 27001 audits, and distributed infrastructure across cloud availability zones and multiple geographic regions. Constructor was founded in 2015 and operates with a remote team spread across 40 countries and 375+ team members.
Why use Constructor?
- Constructor ties multiple discovery touchpoints to one reasoning engine, so shoppers get consistent guidance across the journey.
- Its real-time reinforcement learning means discovery can improve as shoppers interact, instead of relying on static rules.
- The platform is built around behavioral signals and product data, which helps it personalize without depending on generic LLM responses.
- Security and privacy are part of the architecture, with minimal personal data collection and audited controls.
- Constructor's scale signals matter: it serves billions of queries every day and is used by large retail brands.
Who is Constructor for?
- Ecommerce merchandising teams who want search and category pages tuned to shopper intent.
- Digital commerce leaders who need discovery experiences that adapt in real time.
- Product teams who want AI shopping and product-question experiences on-site.
- Retail media teams who need sponsored placements without breaking the shopper experience.
- Data and analytics teams who want insight into ranking, attributes, and experiments.
What are Constructor's key features?
AI Shopping Agent
Answers shopper questions and guides product discovery across onsite experiences and ChatGPT, helping retailers turn conversational intent into product matches.
Product Insights Agent
Surfaces product-level insights for merchandising decisions, using AI Product Insights Agent capabilities to show what shoppers search for and click.
Search & Autosuggest
Improves onsite search and autosuggest with commerce-aware ranking, supporting faster product finding and measurable conversion lifts like 4.14% relative lift in sitewide search conversion.
Browse
Organizes category browsing so shoppers can move through large catalogs more easily, supporting discovery across product grids and navigation paths used by brands like REI and West Elm.
Recommendations
Delivers personalized product recommendations across onsite and offsite touchpoints, including email recommendations and SMS & mobile notifications for repeat visits and higher order value.
Retail Media
Lets retailers monetize discovery surfaces with retail media placements, tying sponsored content into search, browse, and collections experiences without breaking the shopping flow.
Attribute Enrichment
Adds and improves product attributes for better filtering and matching, helping teams keep catalog data usable across Shopify-connected commerce experiences.
Merchandiser Controls
Gives merchandisers controls for ranking, boosting, and tuning results, so teams can adjust search and browse behavior without engineering changes.
What does Constructor integrate with?
- Shopify
- ChatGPT
What are Constructor's use cases?
Merchandising for ecommerce teams
Ecommerce merchandising teams use Constructor to tune search and category pages to shopper intent, using Search & Autosuggest and Merchandiser Controls to surface the right products faster. They can also use Browse and Recommendations to keep high-intent shoppers moving toward conversion instead of bouncing.
On-site shopping help
Product teams use Constructor to add AI shopping and product-question experiences on-site, using AI Shopping Agent and Answer Engine Assistants (Beta) to answer shopper questions in the moment. That helps reduce dead-end searches and gives buyers a clearer path to the right product.
Retail media without friction
Retail media teams use Constructor to place sponsored products without breaking the shopper experience, relying on Retail Media and Merchandiser Controls to keep placements relevant. The result is monetization that still feels like part of the natural discovery flow.
Discovery insights for analysts
Data and analytics teams use Constructor to understand ranking, attributes, and experiments, using Product Insights Agent and Agentic Audit to spot where discovery is helping or hurting performance. They can then use Attribute Enrichment to improve the data behind future merchandising decisions.
How does Constructor work?
- Connect your first commerce source, such as Shopify, and let Constructor ingest catalog, search, and shopper behavior data into the platform.
- Configure Search & Autosuggest and Browse to shape how products appear in search results, category pages, and discovery journeys.
- Add AI Shopping Agent or Answer Engine Assistants (Beta) to handle product questions on-site and guide shoppers toward the right item.
- Use Merchandiser Controls, Recommendations, and Retail Media to adjust ranking, promote sponsored placements, and keep discovery commercially relevant.
- Review Product Insights Agent and Agentic Audit to monitor ranking, attributes, and experiments, then refine Attribute Enrichment for ongoing gains.
Frequently asked questions
What is Constructor?
Constructor is an ecommerce discovery platform for merchandising and digital commerce teams that personalizes search, browse, recommendations, and shopping agents across the storefront. Its Commerce Reasoning Engine uses behavioral clickstream and shopper data with reinforcement learning to improve in real time. Key features include AI Shopping Agent, Product Insights Agent, Search & Autosuggest, and Merchandiser Controls. It powers brands such as Sephora, Under Armour, and REI.
What is Constructor used for? Who is it for?
Constructor is used for AI Shopping Agent, Product Insights Agent, and Search & Autosuggest. It's built for Ecommerce merchandising teams, Digital commerce leaders, and Product teams.
Does Constructor have an API and what does it integrate with?
Constructor doesn't publish a public API. It integrates with Shopify, ChatGPT.
Editor's read
Check whether your team needs retail media placements alongside search and browse. Constructor includes Retail Media, but if sponsored content must stay separate from discovery surfaces, validate the merchandising workflow before rollout.
Explore other Commerce & Ecommerce Agents
