# Freyam M. > Founding Engineer, Realm Labs | AI Security through Observability and Trust Visibility Location: San Francisco Bay Area, United States Profile: https://flows.cv/freyam I build infrastructure for a world where software is no longer deterministic. As a founding engineer at Realm, I specialize in hardening AI deployments by treating model uncertainty as a first-class security challenge. My focus is on moving beyond "move fast and break things" toward operational trust, designing zero-trust pipelines for autonomous agents and adversarial frameworks that treat model hallucinations as identifiable system failures. My background is rooted in making complex machines legible. Whether reverse-engineering cloud vulnerabilities at Oracle, optimizing quantum circuit visualizations for Google’s Cirq, or contributing to LLVM MLIR, I’ve always been drawn to the hidden states where systems actually live. From publishing research on bias loops at CHI ’24 to mapping AI attack surfaces into MITRE frameworks, my work is about taking responsibility for systems we can't fully control by building the tools that make them predictable at scale. ## Work Experience ### Founding Engineer (AI/ML Security) @ Realm Labs Jan 2023 – Present | San Francisco Bay Area ### AI/ML Security Researcher @ AI Vulnerability Database Jan 2023 – Jan 2024 | Seattle, Washington, United States • Published ’Closing the Loop: Embedding Observability in the GenAI Product Lifecycle for Systematic Bias Mitigation’ at GenAICHI (ACM CHI 2024), establishing taxonomy-anchored best practices for bias monitoring. • Connected AVID’s multi-domain risk framework with MITRE ATT&CK and ATLAS—partnering with S. Majumdar, C. Anderson, and J. Novikova—to publish an open reference for practitioners and strengthen collective threat modeling. • Engineered BiasAware, a real-time framework for detecting and remediating dataset biases and security flaws through interactive monitoring dashboards. ### DevOps Engineer and Mentor (Google Summer of Code '23) @ Jenkins Jan 2023 – Jan 2023 • Reduced CI build and deployment cycles by 40% through a Jenkins integration plugin that streamlines pipeline event handling and enables parallel stage execution across agents. • Cut communication overhead by 25% by embedding real-time GitLab build updates into Jenkins dashboards. ### Cloud Security Engineer @ Oracle Jan 2023 – Jan 2023 • Designed a two-phase vulnerability assessment framework leveraging Oracle Cloud Guard and the Vulnerability Scanning Service (VSS), enabling automated detection and prioritization of risks in OCI images. • Developed a heuristic-driven risk scoring model for OCI compute instances, incorporating CVE severity, open port analysis, and CIS benchmarks to improve security posture and remediation prioritization. • Built an automated vulnerability retrieval pipeline integrating Oracle security APIs, streamlining assessments and reducing manual compliance checks. • Engineered a dynamic security sandbox to test and isolate potential threats in OCI environments, ensuring real-time scanning without disrupting production workloads. ### Data Science Researcher @ Precog-at-IIITH Jan 2022 – Jan 2022 • Analyzed COVID-19 vaccine rollout with regression and decision trees to identify accessibility and resource patterns. • Generated insights that guided equitable access strategies and informed policy recommendations based on data. ### Full Stack Engineer and Mentor (Google Summer of Code '22) @ Joomla! Jan 2022 – Jan 2022 • Mentored contributors architecting a context-aware guided-tour engine, cutting onboarding friction and support tickets. • Guided development of a zero-config CLI scaffolder, automating boilerplate and halving extension delivery time. ### Technical Product Manager @ Pradical Jan 2022 – Jan 2022 • Led a team of four junior engineers to architect and deploy microservice frameworks, optimizing latency and scalability for enterprise-level e-marketplaces. • Automated operational workflows for 100,000+ farm-to-form businesses by developing and deploying a centralized platform, Puhas Farms, streamlining resource allocation and supply chain management. • Accelerated yearly trades of over 2 million rupees by creating a contract-based auction space (INBEX). ### Cloud Architect @ ColdEasy (Temperate Technologies) Jan 2022 – Jan 2022 • Optimized server performance to sub-50ms response times by engineering a hybrid cloud infrastructure with Kubernetes and Istio, enhancing data processing efficiency and service reliability. • Drove a 2000% QoQ growth in user base by expanding the product ecosystem with cross-platform dashboards. ### Software Engineer @ Indian School of Business Jan 2021 – Jan 2022 • Increased the participation of semi-literate forest landowners by over 300% by developing a gamified species classifier (SpecieFi). • Simplified the legal processes of the FRCs in over 700 Indian communities by spearheading a localised solution (RatiFi and FormiFi). • Enhanced data security with robust encryption and access controls, safeguarding sensitive information on community resources. ### Project Developer (Cirq, Google Quantum AI) @ Google Jan 2022 – Jan 2022 • Developed visualizations for Matrix Gates, enabling dynamic graphical representation of custom gates, including multi-qubit operations, to improve interpretability and analysis. • Strengthened the Cirq testing framework by implementing comprehensive unit tests for SVG rendering, ensuring reliable visual outputs across diverse quantum circuit configurations. ### Project Developer (MLIR, LLVM) @ LLVM Foundation Jan 2021 – Jan 2021 • Engineered a visualization tool within the MLIR framework to represent Rewrite Rules and Pattern Matching DAGs, enhancing the comprehension of transformation sequences. • Developed a recursive DAG dump feature akin to LLVM's dumpr, enabling graphical differentiation of input/output DAG pairs with color-coded highlights for node modifications and pattern matches. • Integrated graph analysis techniques to track and display the impact of greedy and dialect-specific rewrite drivers, providing deep insights into the optimization process and transformation efficacy. ### Data Science Engineer (Dask, Google Summer of Code '21) @ NumFOCUS Jan 2021 – Jan 2021 • Achieved 600% performance improvement and sub-100ms response time by optimizing Dask collections visualisation. • Played a pivotal role in partnering with NVIDIA and Anaconda to enhance Dask’s parallel computing capabilities. ### Open Source Contributor (OpenMRS, LibreHealth, etc.) @ Google Code-in Jan 2019 – Jan 2019 ### Open Source Contributor (Wildfly by RedHat, CCExtractor, RTEMS etc.) @ Google Code-in Jan 2018 – Jan 2018 ## Education ### B.Tech. (Hons.) in Computer Science and Engineering International Institute of Information Technology Hyderabad (IIITH) ## Contact & Social - LinkedIn: https://linkedin.com/in/freyam-mehta - Website: https://freyam.dev/ --- Source: https://flows.cv/freyam JSON Resume: https://flows.cv/freyam/resume.json Last updated: 2026-04-05