Project Overview We are building a brand-new, unified product search experience inside Adobe Experience Manager (AEM as a Cloud Service). The search experience must combine: Ranked search results from OpenSearch 8.x Product attributes from Plytix PIM (retrieved via AWS Lambda APIs) Content stored in AEM (pages, assets, content fragments) The objective is to create a fast, modern, faceted, multilingual search that supports: A Search Results Page (with filtering, sorting, and pagination) A Product Detail Page (data retrieved from PIM/OpenSearch) Autocomplete and fuzzy search features This will replace our existing static AEM search and prepare the foundation for a global search experience. Current Environment You will be working with the following infrastructure: AEM AEM as a Cloud Service (Adobe Managed) Components exist, but no search integration currently implemented Cloud Manager CI/CD pipeline is in use Multilingual sites exist; must support EN/ES/PT/JA OpenSearch OpenSearch 8.x cluster is provisioned Multiple indices anticipated (content, products, assets) AWS SigV4 authentication already configured PIM (Plytix) Plytix PIM exposes catalog/product data via: AWS Lambda API endpoints JSON responses containing product attributes, specs, and images Internal Implementation (Provided by Our Team) Our team is responsible for: Sending product data from PIM → OpenSearch Maintaining the indexes and pipelines Defining product attribute mappings Your responsibility is AEM-side search integration and UI templates only. Your Mission Deliver a production-ready AEM search integration that enables: AEM components → query OpenSearch → return ranked results For each search hit: Pull product attributes from Plytix PIM API (real-time or cached) Enrich and merge results before rendering Faceting, filtering, and autocomplete Search result components for AEM Author + Publish Product detail page powered by OpenSearch/PIM data Configurable OSGi components for: Endpoint URLs Index names Query tuning Facets/filters Language analyzers Everything must deploy through Cloud Manager and work within AEM’s ACS constraints. Deliverables 1. Architecture & Design Technical architecture diagram showing AEM ↔ OpenSearch ↔ PIM interactions Field mapping between search results and PIM attributes Query tuning recommendations Index structure recommendations (provided jointly with our internal team) Decision document: real-time PIM fetch vs cached enrichment 2. AEM Search Results Page (SRP) Build a complete AEM React/HTL component for: Search bar Results grid/list Pagination Sorting (e.g., Relevance, Part Number) Facets Product type Series Dimensions Region availability Filters Price (if applicable) Size ranges (if applicable) Markets/Industries Must support: Multilingual labels Author & Publish environments Lighthouse performance parity Responsive rendering SEO-safe rendering (SSR preferred) 3. Product Detail Page (PDP) Create an AEM template for product detail: Part number Images Description Product series Key attributes/specifications (from PIM) CAD download links (if provided via API) Breadcrumb Related products (optional / phase-2) Data source: OpenSearch hit + PIM attribute API enrichment. 4. Autocomplete Implementation AEM autocomplete component Uses OpenSearch completion suggester Returns: Part numbers Product series Categories Popular queries Debounced queries required; must be high performance. 5. AEM ↔ OpenSearch Connector Build an AEM OSGi service that: Authenticates with OpenSearch (SigV4) Constructs queries (facets, ranges, fuzziness) Parses and normalizes OpenSearch responses Exposes results to components via Sling Models Supported query capabilities: Prefix search Partial search (ex: “620” matches “6205ZZ”) Fuzzy search Facets & ranges Pagination Relevance tuning 6. Real-Time PIM Enrichment Layer Implement a service that: Calls Plytix PIM’s AWS Lambda endpoint for each product hit Merges PIM attributes with OpenSearch hits Caches results to avoid API overuse (strategy TBD) Handles API failures gracefully 7. Testing & Validation Unit tests (70%+ coverage) Integration tests (mock OpenSearch + mock PIM API) Performance validation (no Lighthouse score drop) Multilingual testing Cross-browser/cross-device QA 8. Deployment & Handoff Deploy to AEM DEV → STAGE → PROD (via Cloud Manager) Code documentation (Javadoc + README) Handoff package: Component diagrams Extension points How to tune relevance How to add new facets How to add new product attributes Acceptance Criteria AEM pages load with no performance regression (Lighthouse reference score provided) Result accuracy matches OpenSearch ranking + PIM data Search works across all languages Autocomplete returns results <100ms on Publish Facets, filters, sorting all functional Product detail page renders full attribute set Code is production-grade, well documented, and test-covered Required Skills Strong experience with AEM as a Cloud Service Sling Models, HTL, OSGi services Cloud Manager Proficiency OpenSearch 7/8.x (or Elasticsearch experience) Experience integrating with external APIs Knowledge of AWS SigV4 (bonus) Performance tuning experience Preferably experience with PIM integration