AI Engineer

Must-Have Skills 2+ years in software development, with 1-2+ years in AI/ML / data-heavy products / ad-tech / mar-tech / e-commerce tools. Strong backend skills in Python (FastAPI/Django/Flask) or Node.js/TypeScript. Practical ML experience: scikit-learn, XGBoost, or deep learning frameworks (PyTorch/TensorFlow). Comfortable working with large datasets, feature engineering, model evaluation. Experience with 3rd party APIs, ideally: Amazon SP-API / Advertising API, or other marketplace/ad APIs. Strong knowledge of SQL and relational databases (PostgreSQL/MySQL). Good understanding of cloud platforms (AWS/GCP/Azure), Docker, task queues (Celery/Resque/RQ, etc.). Ability to own a project end-to-end: architecture implementation deployment iteration. Key Responsibilities1. Product & Architecture (Helium 10–style tool)Design overall system architecture for an AI-powered SaaS tool for: Product & keyword research Competitor tracking Ads & campaign optimization Listing quality & ranking insights Build a scalable, modular backend so we can plug in more marketplaces and ad channels over time. Decide on tech stack, data storage, and cloud architecture (with founder). 2. API Integrations (Amazon + Ads + Analytics)Integrate with platforms such as: Amazon SP-API / Advertising API Google Ads, Meta Ads, other ad platforms (later) Analytics tools if required Build data ingestion services to: Sync products, keywords, campaigns, orders, and performance data Normalise and join data across platforms Handle OAuth, tokens, refresh logic, and rate limits Create reusable connectors so new marketplaces/APIs can be added quickly. 3. AI / Machine Learning ModelsDesign and implement ML/AI models for: Performance forecasting & campaign duration planning Keyword harvesting / keyword recommendations Budget & bid optimization suggestions Audience/placement insights Anomaly detection (sudden drop in ROAS, spike in ACoS, etc.) Experiment with different approaches: classic ML, time-series forecasting, clustering, and (where relevant) LLM-based analysis. Continuously improve models using real campaign data and feedback from marketers. 4. Data Analysis & VisualisationBuild dashboards and visualizations for: Performance by campaign / ad group / keyword Cross-channel view (Google + Meta + Amazon etc.) Lifetime value, ROAS, TACoS, ACOS, profitability, etc. Work with UX/UI or front-end devs to make insights simple, visual, and actionable for non-technical users. 5. Productisation & SaaSTurn models and analytics into SaaS features: Recommendations widgets (e.g., Pause these 3 keywords, Increase budget here) Automated rules / workflows (e.g., trigger alerts or changes based on conditions) Contribute to multi-tenant architecture, billing logic, roles & access, and usage logging. Collaborate with the team on roadmap, feature prioritization, and beta testing with real clients. 6. Quality, Security & DocumentationWrite clean, maintainable, well-tested code. Implement basic MLOps practices: model versioning, monitoring, and performance tracking. Maintain clear technical documentation for APIs, data schemas, and models. Follow best practices for data privacy and security, especially around client ad accounts. 7. SaaS & Multi-tenant PlatformBuild a secure multi-tenant SaaS: User management, roles & permissions Subscription plans, usage limits Billing integration (Stripe/Razorpay/etc.) Implement logging, monitoring, and error tracking to keep the system stable.