Pebble - Senior AI/ML Engineer - Cloud Sustainability & Workload Profiling

About Pebble Falcon :

Falcon is a platform built by Pebble to help enterprises reduce compute waste, optimize infrastructure efficiency, and minimize energy and carbon costs across hybrid cloud environments.

Our AI-powered insights help businesses make smarter decisions about when and how their workloads run driving major savings and sustainability.

Key Responsibilities :

- Analyze real-time cloud emissions/energy data collected from Falcon and generate actionable insights.

- Develop models to forecast emissions trends based on historical workload data and AI-driven anomaly detection.

- Profile AI workloads (training & inference models) to determine energy footprint per model, GPU utilization efficiency, and energy cost per inference run.

- Build carbon-aware workload scheduling models that optimize AI jobs across different cloud regions based on renewable energy availability.

- Develop AI-powered recommendations for reducing cloud emissions and optimizing workloads (e.g., instance right-sizing, model quantization, auto-scaling improvements).

- Leverage OpenTelemetry, eBPF, or custom profiling tools to extract deep insights into how AI models consume compute, memory, and GPU cycles.

- Collaborate with data engineering teams to build scalable pipelines for emissions analytics using Spark, Flink, or Dask.

- Ensure model explainability by developing interpretable AI-driven insights for cloud sustainability reporting.

Required Skills and Qualifications :

- 5+ years of experience in AI/ML engineering, data science, or cloud infrastructure analytics.

- Expertise in AI model profiling, deep learning efficiency, and cloud workload optimization.

- Strong knowledge of PyTorch, TensorFlow, and ONNX for AI workload analysis.

- Experience with GPU profiling tools (NVIDIA Nsight, TensorRT, Triton Inference Server, DeepSeek).

- Strong programming skills in Python, Go, or Rust for building ML pipelines and cloud integrations.

- Experience with time-series forecasting, anomaly detection, and predictive modeling.

- Familiarity with AWS/GCP/Azure cloud APIs to extract resource utilization data for AI workloads.

- Experience with MLOps pipelines and automated model monitoring (Kubeflow, MLflow, SageMaker).

- Understanding of carbon-aware computing, energy-efficient AI, and green computing strategies.

Bonus or Good-to-Have Skills :

- Experience with LLM model efficiency (e.g., pruning, quantization, distillation).

- Knowledge of sustainability reporting frameworks (GHG, SASB, CSRD).

- Familiarity with Green AI research and techniques for reducing model energy consumption.

- Experience with serverless AI inference to optimize cloud energy usage.

Why Join Us ?

- Tackle one of the biggest challenges in AIreducing the carbon impact of machine learning models.

- Build an AI-powered sustainability observability platform for cloud.

- Work at the intersection of AI, cloud infrastructure, and sustainability.

- Join a fast-moving startup with the ambition to change how enterprises approach cloud sustainability.