cv
General Information
Full Name | Bobby Filar |
Location | Minneapolis, MN, USA |
Experience
-
2022 Head of Data Science
Sublime Security - Designed a computer vision feature for the detection and identification of brand impersonation and phishing portals
- Designed a natural language understanding model to perform intent classification and named entity recognition
- Designed malware classifier for detecting malicious macro documents
- Integrated these models into a robust API service, providing seamless access to users with minimal latency
- Conducted rigorous testing to evaluate the effectiveness of the models in detecting cyber threats, achieving high accuracy rates
- Optimized the models to reduce computer resource consumption, enabling them to operate efficiently on millions of emails daily and reducing the cost of computing resources.
-
2019-2022 Senior Manager, Security Data Science
Elastic - Responsible for recruiting and hiring (3x growth in 18 months).
- Sets research agenda by collaborating with cross-functional teams to design product features that leverage machine learning.
- Manages ML projects, including identifying data needs, setting go/no-go performance metrics, and communicating to stakeholders.
- Technical lead of Insights Initiative, an ML-backed approach to improving alert triage experience within the Elastic Security App.
- Focuses research on holistic ML approach (supervised + unsupervised learning) on tailoring detection experience to local environments.
- Researching the use of language models to detect anomalous behavior in event data.
- Built ML service on top of alert feedback data to reduce global false positives by ~40% within 48 hours post-model releases.
-
2015-2019 Director of Data Science
Endgame (acquired by Elastic) - Recruited and led a distributed Data Scientists and Engineers team focused on delivering ML features and critical data pipeline enablers.
- Acted as an independent monitor of ML projects to track progress, identify challenges, and present results to key stakeholders.
- Designed and developed Artemis, a natural language interface for querying security event data.
- Developed ML features for a static PE binary malware classifier.
- Designed an Adversarial ML service for tree-based classifiers to discover blind spots and identify potential decay.
- Designed and implemented an Active Learning interface for increasing NLP model efficacy.
-
2011-2015 Senior Data Scientist
Battelle Memorial Institute - Designed and implemented a social media collection, analysis, and visualization platform.
- Developed an authorship attribution model for source code attribution.
- Built a model to demonstrate information diffusion within small online communities and predict future propagation
Patents
-
2019 Voice and textual interface for closed-domain environment (US20190088254A1)
-
2021 Chatbot interface for network security software application (US20210176282A1)
-
2022 Systems and Methods of Anomalous Pattern Discovery and Mitigation (US20220100857A1)
Education
-
University of Pittsburgh
- Graduate Studies -- Security and Intelligence
-
Ohio University
- Bachelor of Science -- Interational Communcation