Job Description - AI Architect (B2B Product Search | Wholesale Distribution) | 10-15 Years
Role Overview
We are seeking an experienced AI Architect to lead the design and implementation of scalable, enterprise-grade AI solutions with a strong focus on AI-driven B2B Product Search for a Wholesale Distribution enterprise. The role will architect end-to-end solutions spanning search/retrieval, ranking, GenAI augmentation, data pipelines, security, and cloud-native delivery. The ideal candidate combines strategic thinking with hands-on delivery using Python, AKS, MongoDB, and Google Gemini to deliver high-relevance, low-latency product discovery experiences at scale.
Mandatory Domain Experience (B2B Product Search - Wholesale Distribution)
Proven experience building AI/ML models for B2B product search in wholesale distribution, industrial distribution, or adjacent B2B commerce domains (product discovery, product matching, search relevance, ranking).
Demonstrated experience delivering measurable relevance improvements using semantic search, hybrid retrieval, and reranking strategies.
Experience integrating search into enterprise commerce and ordering workflows and modernizing legacy search stacks to support AI and analytics.
Ability to demonstrate business impact (e.g., reduction in zero-result queries, improved search success, improved conversion or assisted-ordering/quoting outcomes).
Core Responsibilities
1) B2B Product Search Architecture (End-to-End)
Design end-to-end AI search pipelines: ingestion enrichment indexing retrieval ranking GenAI augmentation (RAG where applicable).
Architect product discovery and product matching solutions using hybrid retrieval and reranking approaches.
Define technical blueprints for integrating search into wholesale distribution workflows (catalog browsing, order entry, assisted selling, and quoting support).
Establish evaluation strategy for relevance and groundedness, including offline benchmark sets and online KPI instrumentation.
2) Search Relevance, Ranking & B2B Merchandising
Lead relevance optimization using embeddings, semantic similarity, metadata-aware retrieval, and learning-to-rank/reranking strategies.
Enable B2B merchandising patterns such as recommendation rails, bundled products, and top-ordered products to improve discovery and cross-sell.
Implement monitoring for relevance drift, query intent shifts, and catalog quality issues; recommend corrective actions.
Define fallback strategies (lexical, curated results, synonym/taxonomy expansion) to maintain search quality and resiliency.
3) Data Pipeline & Product Knowledge Foundation
Architect scalable pipelines for structured product data and unstructured content (specs, datasheets, manuals, images metadata).
Build enrichment strategies to improve findability (attribute normalization, synonym management, taxonomy/category mapping, unit conversions, spec harmonization).
Design data storage and access patterns using MongoDB and supporting vector/hybrid indexes.
Support customer-, branch-, and entitlement-aware product contexts where applicable (visibility constraints and segment-specific relevance).
4) Cloud-Native AI Platform Delivery
Deploy AI services on AKS with production-grade reliability, CI/CD integration, and observability.
Build API-first search services using Python microservices with secure access controls.
Integrate Google Gemini for GenAI augmentation, tool orchestration, and response grounding.
Ensure cost-aware architecture across embedding generation, retrieval, and inference.
5) Scalability, Performance & Reliability
Ensure low-latency search responses and scalability for high-traffic and large catalog volumes.
Implement performance tuning: caching, async processing patterns, index optimization, and efficient query routing.
Define SLAs/SLOs for search services and implement reliability engineering practices (monitoring, ing, incident response).
6) Collaboration & Leadership
Work closely with Product, Engineering, Data Science, and DevOps teams to translate business search requirements into technical architecture.
Provide technical leadership, mentor teams, and drive adoption of modern AI search practices.
Create architecture artifacts (HLD/LLD, reference architectures, API contracts, data contracts) and guide implementation reviews.
7) AI Governance & Responsible AI
Establish standards for Responsible AI, privacy, security, and compliance in search outputs.
Ensure solutions are auditable and secure, including access control, data handling, and prompt/retrieval safety.
Define governance for model and prompt lifecycle (versioning, evaluation gates, monitoring, rollback).
Tech Stack (Must Have)
Python (mandatory)
AKS (Azure Kubernetes Service) for deployment and scaling
MongoDB (NoSQL storage for product/search data patterns)
Google Gemini (LLM integration for GenAI search augmentation / assistants)
Good to have:
Vector databases (e.g., Pinecone, Milvus, Weaviate)
Search engines (e.g., Elasticsearch, OpenSearch)
Evaluation frameworks for relevance/groundedness and A/B testing
MLOps practices (model registry, feature store, model monitoring)
Experience & Qualifications
10-15+ years in software engineering, data engineering, or architecture roles.
5+ years building AI/ML systems with recent exposure to GenAI/LLMs.
Proven enterprise experience designing and implementing AI-driven solutions end-to-end.
Bachelor s or Master s degree in Computer Science, Data Science, or a related field.
Leadership & Communication
Ability to map B2B business challenges (e.g., low search success, low search-to-buy, quoting friction) to AI-driven technical solutions.
Strong stakeholder communication: explain complex AI search architecture to non-technical audiences.
Balance strategic thinking with hands-on execution and pragmatic delivery.
Success Metrics
Improved product findability and relevance for B2B catalog search (reduced zero-results, improved search success).
Increased engagement and revenue impact through AI-driven discovery patterns (recommendations, bundles/top-ordered) and improved content quality.
Improved platform performance, scalability, reliability, and operational excellence for search services deployed on AKS.
