Full-Stack AI Engineer

Job Title: Full-Stack AI Engineer Position Type: Full-Time, Remote Working Hours: U.S. client business hours (with flexibility for model deployments, experimentation cycles, and sprint schedules)About the Role: Our client is seeking a Full-Stack AI Engineer to design, build, and deploy AI-powered applications. This role requires bridging software engineering with applied machine learning, ensuring that models are integrated into production systems that are scalable, reliable, and user-friendly. The Full-Stack AI Engineer combines back-end services, front-end interfaces, and machine learning pipelines to deliver practical, business-driven AI solutions.Responsibilities:AI Model Integration:Deploy pre-trained and fine-tuned ML/LLM models (OpenAI, Hugging Face, TensorFlow, PyTorch).Wrap models in APIs (FastAPI, Flask, Node.js) for scalable inference.Implement vector search integrations (Pinecone, Weaviate, FAISS) for retrieval-augmented generation (RAG).Data Engineering & Pipelines:Build ETL pipelines for ingesting, cleaning, and transforming text, image, or structured data.Automate data labeling, preprocessing, and versioning with Airflow, Prefect, or Dagster.Store and manage datasets in cloud warehouses (Snowflake, BigQuery, Redshift).Application Development (Full-Stack):Build front-end UIs in React, Next.js, or Vue to surface AI-powered features (chatbots, dashboards, analytics).Design back-end services and microservices to connect models to business logic.Ensure responsive, intuitive, and secure interfaces for end users.Infrastructure & Deployment:Containerize ML services with Docker and deploy to Kubernetes clusters.Automate CI/CD pipelines for model updates and application releases.Monitor latency, cost, and model drift with MLflow, Weights & Biases, or custom dashboards.Security & Compliance:Ensure AI systems comply with data privacy standards (GDPR, HIPAA, SOC 2).Implement rate limiting, access control, and secure API endpoints.Collaboration & Iteration:Work with data scientists to productionize prototypes.Partner with product teams to scope AI features aligned with business needs.Document systems for reproducibility and knowledge transfer.What Makes You a Perfect Fit:Strong coder with a foundation in both full-stack development and applied ML/AI.Comfortable building prototypes and scaling them to production-grade systems.Analytical problem solver who balances performance, cost, and usability.Curious and adaptable, staying current with emerging AI/LLM tools and frameworks.Required Experience & Skills (Minimum):3+ years in software engineering with exposure to AI/ML.Proficiency in Python (PyTorch, TensorFlow) and JavaScript/TypeScript (React, Node.js).Experience deploying ML models into production systems.Strong SQL and experience with cloud data warehouses.Ideal Experience & Skills:Built and scaled AI-powered SaaS products.Experience with LLM fine-tuning, embeddings, and RAG pipelines.Knowledge of MLOps practices (Kubeflow, MLflow, Vertex AI, SageMaker).Familiarity with microservices, serverless architectures, and cost-optimized inference.What Does a Typical Day Look Like? A Full-Stack AI Engineer's day revolves around connecting models to real-world applications. You will:Review and refine model APIs, testing latency and accuracy.Write front-end code to surface AI features in user-friendly interfaces.Maintain pipelines that clean and prepare new datasets for training or fine-tuning.Deploy updates through CI/CD pipelines, monitoring cost and performance post-release.Collaborate with product and data science teams to prioritize AI features that solve real user problems.Document workflows and results so solutions are repeatable and scalable. In essence: you ensure AI moves from prototype to production — reliable, compliant, and impactful.Key Metrics for Success (KPIs):Successful deployment of AI features to production on schedule.Application uptime ≥ 99.9% and inference latency < 500ms for key endpoints.Reduction in manual workflows replaced by AI features.Model performance tracked and stable (accuracy, drift, false positives/negatives).Positive user adoption and satisfaction of AI-driven features.Interview Process:Initial Phone ScreenVideo Interview with Pavago (https://himalayas.app/companies/pavago) RecruiterTechnical Assessment (e.g., deploy a small ML model with API endpoints and basic front-end integration)Client Interview(s) with Engineering TeamOffer & Background Verification Originally posted on Himalayas (https://himalayas.app)