Lead AI Engineer

Role Overview

We are hiring a Lead AI Engineer to provide hands-on technical leadership for AI systems powering ecommerce search, retrieval, ranking, personalization, and visual discovery. This is a senior individual-contributor role centered on research, rapid experimentation, production engineering, and measurable business impact across text, image, and multimodal experiences.

You will set technical direction and lead end-to-end AI solutions across matching, multimodal retrieval, ranking, computer vision, vision-language modeling, and GenAI-powered experiencesfrom problem framing and research through deployment, evaluation, and iteration at scale. You will also mentor engineers and scientists, establish reusable technical patterns, and influence cross-functional product and platform roadmaps.

Core Responsibilities

• Set technical direction and lead research, experimentation, and production delivery for ML models spanning:
• Product matching and entity resolution
• Semantic, visual, and multimodal retrieval
• Search relevance, ranking, re-ranking, and recommendation
• Image classification, detection, segmentation, similarity, visual attribute extraction, and content understanding
• Architect and optimize multi-stage retrieval and ranking pipelines across text, image, and fused multimodal representations.
• Develop, fine-tune, and evaluate computer vision, vision-language, and multimodal models, including image/text encoders, multimodal embeddings, contrastive learning approaches, and vision-language foundation models.
• Create rigorous offline evaluation, benchmarking, error analysis, and online experimentation frameworks; define success metrics and translate model improvements into customer and business outcomes.
• Apply reinforcement learning, learning-to-rank, or bandit methods where appropriate for personalization and ranking optimization.
• Build and deploy GenAI and agentic AI systems that reason across text and visual content for search, discovery, catalog, and content use cases.
• Lead model and data strategy for multimodal systems, including training data quality, annotation, weak supervision, hard-negative mining, robustness, bias, safety, and responsible use of visual data.
• Productionize ML systems using Databricks-based pipelines and AKS, designing for latency, throughput, reliability, observability, and cost efficiency.
• Establish reusable architectures, evaluation standards, and MLOps/LLMOps practices that improve delivery quality and engineering velocity across the team.
• Independently frame ambiguous, high-impact problems and drive them from research hypothesis through production adoption.
• Provide hands-on technical leadership through design reviews, mentoring, technical decision-making, and collaboration with Product, Engineering, Data, and business stakeholders.

Required Qualifications

• 6+ years of experience in machine learning, applied science, ML engineering, or applied AI, including experience leading technically complex initiatives from research through production.
• 2+ years of hands-on experience developing, training, fine-tuning, and evaluating computer vision models for production or applied research use cases.
• Deep expertise in at least two of the following, with the ability to work across the broader portfolio:
• Search, information retrieval, learning-to-rank, recommendation, or entity matching
• Computer vision and deep visual representation learning
• Vision-language models, multimodal embeddings, or multimodal foundation models
• Hands-on experience designing and evaluating image, text, and multimodal retrieval or ranking systems, including relevant datasets, loss functions, sampling strategies, benchmarks, and error analysis.
• Hands-on experience developing and evaluating ranking and re-ranking models, including learning-to-rank, neural or transformer-based rankers, cross-encoders, feature-based models, hard-negative mining, and offline and online relevance evaluation.
• Strong knowledge of modern deep learning architectures and training or adaptation techniques, including transformers, contrastive learning, transfer learning, fine-tuning, and embedding-based methods.
• Proven track record of building and operating production-grade ML systems with measurable product or business impact.
• Strong proficiency in Python and PyTorch or an equivalent deep learning framework.
• Hands-on experience with Databricks, distributed data/model pipelines, and Kubernetes-based deployment; AKS experience is preferred.
• Demonstrated technical leadership: setting direction, making architecture decisions, raising scientific and engineering standards, mentoring others, and influencing cross-functional stakeholders without relying on formal authority.
• Ability to communicate complex research and system trade-offs clearly to technical, product, and business audiences.

Preferred Qualifications

• Master’s or PhD in Computer Science, Machine Learning, Computer Vision, NLP, Information Retrieval, or a related field, or equivalent practical experience.
• Experience with:
• Vision transformers, image-text models, multimodal large language models, or models such as CLIP-like dual encoders
• Visual search, image similarity, multimodal product matching, OCR/document understanding, or product attribute extraction
• Large-scale approximate nearest-neighbor retrieval, vector databases, multi-stage ranking, and recommendation systems
• Model adaptation, distillation, quantization, or efficient multimodal inference
• GenAI, LLMs, retrieval-augmented generation, and agentic frameworks
• Ecommerce, marketplace, catalog, or digital merchandising domains
• Familiarity with:
• Modern data platforms, data lakes, and scalable GPU training or inference environments
• MLOps/LLMOps tools such as MLflow
• AI governance, evaluation, explainability, safety, privacy, and responsible use of image and multimodal data
• Evidence of technical thought leadership through patents, publications, open-source contributions, internal platform leadership, or repeated delivery of novel AI capabilities is a plus.