16 days ago
Budapest, HungarySenior
Responsibilities
- Own detection capabilities from problem framing and data preparation through experimentation, production deployment, and customer-impact telemetry.
- Build and maintain data pipelines for sample collection, labeling, ground-truth management, feature extraction, and versioned training and evaluation datasets.
- Design, train, fine-tune, evaluate, and deploy models as versioned runtime-portable artifacts with documented specifications, benchmarks, and limitations.
- Build automated model training and release pipelines with reproducible runs, evaluation gates, and versioned artifact publishing.
- Monitor production model quality through telemetry feedback loops, false-positive escalations, drift monitoring, and retraining cadence.
- Design evaluation methodologies for adversarial and distribution-shifting domains, including time-split validation and robustness against evasion.
- Use AI-assisted development and build agentic automation for labeling assistance, evaluation, regression testing, and reporting.
- Set an experimentation roadmap focused on measurable customer-facing detection gains and share findings with the team.
Requirements
- At least 3 years of applied machine learning experience across at least two relevant areas, including text or content classification, computer vision, similarity search or embedding models, and security or threat detection.
- Demonstrated end-to-end delivery of personally developed models from data through production systems used by teams or customers.
- Experience building model-training data pipelines, including collection, labeling, ground-truth management, feature extraction, and dataset versioning.
- Strong Python skills and experience with PyTorch or TensorFlow and model evaluation frameworks.
- Strong foundation in statistics and experiment design, including evaluation under distribution shift.
- Ability to work across model integration boundaries, such as ONNX consumption from JVM services.
- Fluency with AI coding and agent tooling and experience automating personal workflows with these tools.
- Excellent communication and collaboration skills and fluency in English.
Benefits
- Opportunity to work on innovative cybersecurity and critical-infrastructure protection technology.
- Collaborative culture with opportunities to work alongside experienced cybersecurity professionals.
- Ongoing training and development opportunities for professional growth.
- Mission-driven work protecting organizations, data, and digital assets from evolving cyber threats.
- Team-building activities and social events in a fun, dynamic atmosphere.
- Role based in OPSWAT’s newly opening Budapest office.
Categories
Data EngineeringML Engineering
About OPSWAT
OPSWAT builds cybersecurity software for critical infrastructure, delivering malware prevention, content disarm and reconstruction, secure file transfer, and device access control across IT, OT, and ICS environments. It sells enterprise software and SaaS platforms to public- and private-sector organizations worldwide. Founded in 2002 and headquartered in Tampa, Florida, the privately held company is used in regulated industries such as energy, government, and finance.