Publications & Writing
Peer-Reviewed
Saluki: Finding Taint-style Vulnerabilities with Static Property
Checking
BAR/NDSS Workshop 2018. Research tool for detecting taint-style
security properties in binary code using a mix of static and
dynamic taint techniques. Provides a DSL for taint policies.
Demonstrated on real programs.
Role: Author/maintainer of repo. Links: GitHub repo Paper
DeCoCo: Blockchain-based Decentralized Compensation of Digital
Content Purchases
IEEE BRAINS 2020. Blockchain-based decentralized compensation
for digital content purchases with transparent, programmable
payouts and smart-contract design.
Role: Author/co-author. DOI: 10.1109/BRAINS49436.2020.9223299 ResearchGate
Articles & Guides
Utility NFTs more than simple JPEGs?
Medium, Feb 1, 2023. Defines "utility NFTs" and argues for
real-world utility beyond collectibles, with examples like
access passes. Implications for marketplaces.
Link: Medium Article
How to run your python code in KnowledgeX
Medium, Feb 17, 2022. Guide on running Python jobs on KnowledgeX
via iExec and TEEs. Covers Dockerized templates, worker
verification, and auditability for developers and data
scientists.
Link: Blog Post
REL_2020_02_APP_web.pdf
Placeholder: Title and abstract pending confirmation.
[TODO: Confirm title/abstract from source].
Technical Notes
Occasional technical notes and insights from working with AI
systems, distributed architectures, and enterprise software. These
are less formal than talks but capture important lessons and
observations.
Building a prompt optimization system that scales across teams and
maintains consistency in AI outputs. Lessons learned from
implementing a "Prompt Wizard" for enterprise use cases.
Prompt Engineering AI Governance Enterprise
Designing resilient AI workflows using event-driven architecture
patterns. How SQS/SNS can help manage AI pipeline backpressure and
ensure reliable processing.
Event-Driven AI Pipelines AWS
Intelligent routing strategies for multi-model AI deployments.
Cost optimization while maintaining performance and quality across
different model providers.
Multi-Model Cost Optimization Architecture
Essential security considerations for AI platforms: data privacy,
access control, audit trails, and compliance automation in
AI-first applications.
AI Security Compliance Privacy
These publications and notes represent ongoing learning and
observations from building AI systems. For more structured content,
check out the talks and projects sections.