This is a 100% hands-on individual contributor role where you’ll build the AI engines behind our platform—automated image processing, generative content creation, intelligent workflows, and large-scale ML pipelines. You’ll work across computer vision, generative models, automation, and ML infrastructure to deliver production-ready AI systems.
- Computer Vision & Image Understanding
● Product image analysis, object detection, segmentation
● Automated background removal, image enhancement, preprocessing
● Classification, attribute extraction, and visual search systems
● Quality assessment and edge-case detection models
● Depth estimation and scene understanding from 2D images
● Real-time object detection for AR try-on
● Multi-view image analysis and camera pose estimation - Generative AI & Content Creation
● Fine-tune generative models for visual and marketing asset creation
● Text-to-image and image-to-image model pipelines
● AI-generated product descriptions, tags, and metadata
● Work with diffusion models, GANs, transformers
● Texture generation, style transfer, image editing tools
● Synthetic data generation pipelines
● Experiment with the latest foundation models and diffusion techniques - Intelligent Automation & ML Systems
● End-to-end automation for large-scale product catalog processing
● Recommendation and personalization models
● Automated workflows for QC, moderation, and validation
● Predictive models for engagement and conversion
● Anomaly detection and platform monitoring
● Continuous learning and self-improving systems - Production ML Infrastructure
● Deploy/optimize ML models on AWS
● Build scalable inference pipelines (low latency, high throughput)
● Implement model versioning, A/B testing, CI/CD for ML
● Data pipelines for annotation, augmentation, and quality control
● Optimize models for speed, efficiency, and cost
● Monitoring systems for drift, quality, and performance
● APIs/microservices for ML model serving - Research & Innovation
● Explore latest AI/ML trends and cutting-edge models
● Prototype quickly with state-of-the-art models (GPT-4V, Diffusion, SAM, etc.)
● Integrate open-source tools into our production stack
● Run feasibility experiments and contribute to model architecture decisions
● Document learnings and share insights internally - Technical Stack: – AI/ML Frameworks
● PyTorch, TensorFlow, Hugging Face
● OpenCV, YOLO, Detectron2
● Stable Diffusion, ControlNet, Diffusers
● scikit-learn, XGBoost - Deployment & Infrastructure
● FastAPI, ONNX, TorchScript, TensorRT
● AWS (SageMaker, Lambda, EC2, S3)
● Docker, Kubernetes
● PostgreSQL, Redis, MongoDB, Pinecone - Languages & APIs
● Python (primary), JavaScript/Node.js (working knowledge)
● REST, GraphQL, WebSocket - Nice to Have (3D/Graphics)
● Understanding of rendering pipelines
● Familiarity with glTF/USDZ
● Experience with Three.js or Unity/Unreal - Must-Haves
● 5-8+ years experience in AI/ML, strong computer vision background
● Deep expertise in PyTorch/TensorFlow
● Production ML deployment experience
● Strong understanding of CNNs, transformers, detection, segmentation
● Hands-on experience with diffusion models or GANs
● Strong Python skills and ML system design
● Cloud experience (AWS/GCP/Azure)
● Proven record of shipping ML products
● Passion for experimenting with new AI models - Highly Desirable
● Experience in e-commerce/retail imaging or content pipelines
● Background in automation and intelligent workflow systems
● Recommendation/personalization experience
● Familiarity with multimodal models (vision + language)
● Experience with neural rendering or 3D generation
● Open-source contributions or research publications
● Real-time inference optimization
● Strong MLOps understanding
● Ability to build end-to-end ML-driven features
Problems You’ll Solve:
● Automating large-scale product image processing
● Generating high-quality product visuals at scale
● Extracting structured attributes from image datasets
● Reducing manual processes with intelligent automation
● Optimizing inference speed and cost
● Personalizing user experiences with ML
● Monitoring and evaluating models reliably in production