Job Description

Company: Healthcare company

Position: AI/ML Engineer

Location: Remote

Duration: 12-month contract + extensions

Interview Process: first round will be 30 minute interview with tech lead, second round will be 1 hour with HM and another lead and will be technical

Start Date: ASAP

Minimum qualifications

• 6+ years hands-on ML engineering (with significant ML experience prior to 2023): you ve shipped multiple ML systems to production. *Not looking for someone who started working on AI after the pre-trained models/ChatGPT were released*
• Demonstrated production agent build (at least one end-to-end agentic framework delivered to users).
• Strong with classical ML: feature engineering, cross-validation, calibration, regularization, class imbalance, interpretability (SHAP/LIME), time-series (forecasting, seasonality, drift).
• Solid deep learning foundations (CNN/RNN/Transformers), and practical fine-tuning experience (e.g., LoRA/QLoRA, instruction tuning, RAG).
• Proven MLOps: model registry/experiment tracking (MLflow or equivalent), model serving (FastAPI/TF-Serving/TorchServe/TGI/vLLM), observability.
• Fluency in Python and the ML stack (NumPy/Pandas, scikit-learn, XGBoost/LightGBM, PyTorch/TensorFlow).
• Excellent communication; can drive projects independently as an Individual Contributor.

Must have

• Experience with agent frameworks (LangGraph/LangChain Agents, AutoGen, Google ADK) and tool use (function calling, tool routing, planners).
• Retrieval/RAG design: chunking strategies, embedding models, vector stores (FAISS, Pinecone, Weaviate), hybrid search, evals.
• Must have implemented projects involving some classical ML problems(classification, clustering, regression, anomaly detection, time series etc.,)
• A portfolio of production ML systems
• A shipped agentic system used by real users.
• Comfort acting as a self-directed Individual contributor who can turn vision into shipped systems

Responsibilities

• Own end-to-end ML/AI projects: problem framing, data pipelines, modeling, offline/online evals, deployment, monitoring, and iteration.
• Build and productionize agentic workflows (tool-using/multi-step agents with retrieval, planning, and human-in-the-loop), including safety/guardrails and reliability.
• Train classical ML models (tree ensembles, linear models, anomaly detectors, time-series forecasting) and deep learning models when appropriate.
• Operationalize models with CI/CD, feature stores, reproducible training, and model registries
• Monitor and improve live systems: data & concept drift detection, performance regression, bias/fairness, cost/latency; drive remediation playbooks.
• Partner cross-functionally with product, data, and platform teams; write clear design docs
• Work as an Individual contributor with minimal directions

Job Tags

Similar Jobs