SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: PLN 350,700 - 474,400 / annual
Graphcore is seeking a Data Scientist to transform large-scale infrastructure and hardware-level telemetry into actionable insights and predictive intelligence. You will work at the intersection of data science, distributed systems, and infrastructure observability, analyzing data across the full stack—from data center infrastructure down to chip-level signals (power, thermals, performance counters, reliability indicators).
Key responsibilities include analyzing large-scale telemetry datasets across multiple layers: data center (clusters, networking, cooling), system (servers, accelerators), and silicon-level (on-chip sensors, performance counters, voltage, thermal, error signals). You will develop anomaly detection models for infrastructure-level events and hardware anomalies such as thermal hotspots and voltage instability, and build predictive models for fault detection and failure forecasting including hardware degradation and thermal issues.
You will apply statistical and machine learning techniques to identify patterns and root causes, design automated remediation strategies informed by system and hardware-level signals, and collaborate with engineering teams to integrate models into observability and control systems. Working closely with telemetry platform engineers, you will enable closed-loop optimization using real-time hardware telemetry streams and help define schemas, feature extraction pipelines, and aggregation strategies to ensure data quality and usability.
Graphcore is part of the SoftBank Group and is building the future of AI compute with a team of semiconductor, software, and AI experts creating the complete AI compute stack from silicon and software to infrastructure at datacenter scale. The role is hybrid, requiring 2-3 days per week in the Gdańsk office.
