Work Experience
AI-Driven Security Automation
As a key member of the AI engineering team, I contributed to a next-generation platform designed to automate B2B integrations with Cyware's core Threat Intelligence (CTI) and SOAR (Security Orchestration, Automation, and Response) services. This system intelligently parses diverse, unstructured API specifications (like OpenAPI and Postman files) for over 1,000 unique endpoints.
My primary role was engineering the system's core intelligence, which utilized Generative AI and Natural Language Processing (NLP) models. I developed the logic for a Large Language Model (LLM) to semantically understand API schemas, map complex data fields, and autonomously generate functional, sandboxed Python connector code. This solution was packaged into an intuitive React front-end, successfully automating over 95% of the manual developer effort previously required for new client integrations.
To ensure long-term viability, I also engineered an adaptive machine learning model that implemented Reinforcement Learning from Human Feedback (RLHF). This model iteratively refines connector accuracy by learning from user-submitted corrections and historical data structures, creating a powerful, self-improving automation pipeline.
Zero-Trust & Identity Management (IAM)
As an engineer on the core BlockID development team, I analyzed and enhanced the security posture of the company's flagship Multi-Factor Authentication (MFA) and decentralized identity platform, a system that secures over 20 major enterprise clients and 5,000+ users.
My primary project was the research and implementation of a Zero-Trust security model for a new client-facing service. I designed and deployed new context-aware access policies that evaluated user behavior, device posture (via device fingerprinting), and geographic location. This successfully hardened the authentication flow against sophisticated attacks like session hijacking and lateral movement, moving the platform from a static to a dynamic security model.
I also conducted vulnerability analysis and penetration testing on existing authentication endpoints, identified and documented attack vectors in line with the OWASP Top 10, and assisted senior engineers in patching and deploying secure code, contributing to a 40% increase in user verification speed.
Enterprise Full-Stack & AI Integration
I contributed as a full-stack developer on an enterprise data portal using a Java Spring Boot backend (in Eclipse) and a React frontend (in VS Code). This internal platform enabled clients to navigate, search, and manage thousands of complex records from a large-scale, distributed PostgreSQL database.
My standout contribution was designing and implementing the project's AI-powered text correction service. I developed a Python/Flask microservice that integrated a pre-trained Transformer-based NLP model. This service autonomously parsed, validated, and corrected user-input errors in real-time before database submission, which dramatically improved data integrity and the overall user experience.
On the frontend, I developed several key components, including a dynamic, multi-field search bar and interactive data tables. I also refactored and optimized backend API endpoints, analyzing SQL query performance to improve data retrieval speeds by 25%.
Secure Full-Stack Mobile Development
As a lead developer for this mental health awareness organization, I architected and co-developed a cross-platform mobile application using React Native and Expo. The app was designed from the ground up to be a secure, anonymous support hub for users in need.
I personally engineered the secure backend API using Node.js, Express, and a PostgreSQL database, which enabled all core functionalities. This included an anonymous specialist booking system, an encrypted user journaling feature (using AES-256 encryption), and a real-time emergency contact notification service.
My primary focus was security and user trust. I implemented JWT-based authentication for secure sessions and end-to-end encryption for all user data in transit (with SSL) and at rest, ensuring a robust, private, and safe environment for all users.
Projects
- Designed and programmed C++ firmware for an autonomous robot, integrating ultrasonic and IR sensors for robust, real-time obstacle avoidance.
- Implemented actuator control logic for differential drive locomotion and a relay-switched fan for a dual-function navigation and vacuuming system.