Senior AI/ML Engineer (Remote, EST, Anywhere in Pakistan, USD Salary)

Requirements: Strong experience with FastAPI (or equivalent async frameworks), including dependency injection, UV, Pydantic, and async/await patterns (including thread pool executors for blocking operations). Solid understanding of REST API design, including multi-tenancy, pagination, filtering, JWT/OAuth2 authentication, and structured error handling. Proficiency in SQLAlchemy (including async sessions), raw parameterized queries, schema design, and migrations. Hands-on experience integrating multiple LLM providers (e.g., OpenAI, Anthropic, AWS Bedrock, Ollama, Google Gemini, Snowflake Cortex) using provider abstraction layers. Experience with JSON response validation, markdown/code-block extraction, and fallback error handling (preferably using frameworks like Pydantic). Knowledge of prompt engineering techniques, including context injection, temperature/token tuning, and confidence scoring. Familiarity with embedding-based retrieval and similarity scoring. Experience with production-grade agentic frameworks such as Pydantic AI (structured output generation, agents). Strong experience with gradient boosting models (e.g., XGBoost, LightGBM), including GPU-accelerated training, hyperparameter tuning, and evaluation. Expertise in segmentation, anomaly detection, and feature engineering on high-frequency sensor data. Experience with train/test splits, feature engineering, model evaluation (R², MAE, etc.), and experiment tracking (e.g., MLflow). Understanding of when to combine classical ML with LLM-based components (e.g., LLM-assisted labeling, embedding features in tree models). Strong database knowledge, including complex schemas, JSONB, partitioned tables, row-level security, query optimization, and vector extensions (e.g., pgvector). Familiarity with NoSQL databases like MongoDB and specialized databases such as Redis and Qdrant is a plus. Experience with Snowflake (including Snowpark, Model Registry, and Cortex) or equivalent platforms. Hands-on experience with AWS services such as Bedrock, ECS, and EC2. Experience with Docker and CI/CD pipelines. Familiarity with S3 or equivalent object storage solutions. Ability to work within VPN-gated infrastructure. Experience across multiple client environments or industries (consulting background preferred). Exposure to Industrial IoT or sensor data (high-frequency telemetry, signal processing). Experience in NL-to-SQL or text-to-query system design. Ability to handle multilingual data and implement internationalization. Responsibilities: Design and integrate LLM-powered features, including conversational interfaces, AI agents, structured generation, and retrieval-augmented systems. Build and maintain ML pipelines for prediction, anomaly detection, classification, and time-series analysis. Develop backend APIs and services connecting data sources, models, and client-facing applications. Work with structured and unstructured data across relational databases, data warehouses, and external APIs. Optimize model performance and scalability for production environments, including monitoring and fine-tuning. Collaborate with cross-functional teams (product, data, and engineering) to translate business requirements into technical solutions. Ensure code quality, documentation, and best practices for deployment, testing, and maintainability.