Alex Z.

- Designed and built multi-agent runtime architecture with orchestrator-based routing, shadow/canary deployment paths, and support for multiple agent frameworks including Deep Agent, Strands Agents, and Claude Agent SDK. - Evaluated and integrated agent frameworks for production workloads - balancing trade-offs between multi-model flexibility, native tool orchestration, cold start latency, and managed infrastructure cost profiles. - Led org-wide LLM model governance: multi-tier fallback structures, virtual model config validation in CI/CD, and automated deprecation detection across repositories. - Designed LLM observability stack: OpenTelemetry instrumentation, trace pipelines, PII redaction, eval workflows, per-team cost attribution, and anomaly detection for token spend and latency degradation. - Owned AI gateway security posture: zero-trust architecture, least-privilege RBAC, YAML-only config enforcement, multi-layer guardrail pipelines (prompt injection detection, content moderation, secrets scanning), regional data residency compliance, and customer-facing security architecture documentation. - Built a centralized AI Gateway SDK for unified LLM routing with provider failover, region-scoped access, virtual model abstraction, rate limiting, and cost controls. Drove production readiness through automated testing, gradual rollout via feature flags, and monitoring/alerting integration.

-- High-scale distributed platform delivering real-time recommendations and personalized content to over 100M global users. • Developed a recommendation engine using collaborative filtering and vector embeddings to suggest compatible and complementary items based on user interaction patterns. • Built real-time similarity search with Redis Cluster and external API integrations, supporting over 20M daily requests with sub-50ms response times. • Designed an event streaming pipeline with Apache Kafka to process browsing activity, cart modifications, and configuration changes, enabling low-latency updates to recommendation models. • Created a management console for business teams to define recommendation rules, seasonal campaigns, and bundled offers, allowing rapid adjustments without code changes. • Implemented an A/B testing framework with statistical analysis and feature flags to validate recommendation strategies and interface changes. • Improved personalized content delivery through React SSR with dynamic rendering based on device type, user location, and interaction context, increasing engagement metrics by 2.5x in target regions.

-- AI-powered unified data search and summarization platform across multiple enterprise data sources. • Led development of AI-powered microservices platform for unified search and summarization across 30+ enterprise tools (GitHub, Bitbucket, Slack, Gmail, Jira, Confluence). • Developed AI search engine to improve relevance and accuracy of cross-source queries. Designed and fine-tuned NLP models for semantic summarization of multi-source data. • Directed full SDLC, defining tech vision, architecture, and delivery roadmap aligned with business and GTM goals. • Established and enforced engineering best practices: code reviews, testing standards (Jest), CI/CD pipelines (GitHub Actions, Terraform, Kubernetes), and observability tools. • Developed analytics dashboard and tracing layer with built-in logging, alerting, and query path visualization across services. • Recruited and led engineering team; implemented Agile squads and shared OKRs - achieved 90%+ sprint completion benchmarks.

-- Cybersecurity platform for SAST, SCA, and IaC scanning, policy enforcement, and risk analysis across the SDLC. • Designed and developed Analytics & Reporting and Policy Compliance modules across On-Prem, Private Cloud, and SaaS environments in the Checkmarx One platform - used by customers during security audits and contributed to $10M+ in upsell revenue. • Migrated legacy reporting infrastructure to Kubernetes with decoupled microservices and autoscaling - reduced average report generation time by 60% and enabled 10x data volume throughput. • Built Scan Statistical Dashboards and Security Reports used by 1,000+ enterprise customers to track scan coverage, detect policy violations, and assess compliance posture. • Led a cross-functional team to integrate Checkmarx Go (SAST, SCA scan engines) into the unified Checkmarx One platform; contributed to a shared architecture and improved scanning workflows across Checkmarx products.

-- HILTI's interactive platform for managing fire safety projects, asset tracking, and compliance. • Designed, implemented, and maintained scalable RESTful APIs for HILTI’s core services platform - used by internal tools, customer workflows, and partner integrations. • Integrated multiple external and internal services incl. identity management, billing, customer data, and provisioning pipelines - supported onboarding of 10,000+ new customers monthly. • Developed event-driven workflows using RabbitMQ - improved asynchronous processing and reduced API response latency by 30%. • Led technical discovery and requirements with Product, Customer Success, Support, and UX teams to align solutions with SLAs.

-- IoT supply delivery platform connecting household appliances with automated logistics and real-time tracking. • Designed and built high-performance IoT Smart Home platform for supply delivery management to household appliances using AWS Cloud and microservices architecture.

• Developed scalable backend services on EC2, Lambda, and S3 to optimize system throughput and availability. • Implemented IoT protocols MQTT and CoAP - reduced system latency by 20% and improved responsiveness by 15%.

Led research in EU deep learning & computer vision program - partnership with academic teams across the U.S., EU, and China.

- Ph.D. Internship - Research group led by Prof. Lorenzo Bruzzone: https://scholar.google.com/citations?hl=en&user=ff9-TK4AAAAJ - Research Focus: Principal Component Analysis for the Compression of Hyperspectral Data

Research Focus: Multichannel Remote Sensing Data (Denoising, Classification, Compression, Restoration)