Ben Nguyen

Logo

Bilingual Data Scientist

View My GitHub Profile

About Me:

I am Ben, I am a freelance data scientist on Upwork with experience in the private sector, non-profits, and the Canadian government. I deliver data-driven AI and engineering solutions that strengthen business strategy and generate measurable results.

Toastmaster:

I am a member of Toastmasters International since 2019; I currently belong to the McGill University Toastmasters Club. My journey has taken me to virtual meetings across Ottawa, Montreal, Quebec, Washington D.C., Geneva, and Japan. Toastmaster has helped by developing a strong command of public speaking, leadership, and presentation. It has improved my clarity, my confidence, and my adaptability when presenting complex ideas to different audiences.

Projects:

DriveSmart: An Autonomous Mobility Solution:

Description:

This project delivers a Bluetooth-enabled application for secure, remote activation of an autonomous vehicle, verify the user with encrypted biometric data via PySyft (≥98% accuracy), detect faults using PyTorch (95% precision), show diagnostics on Plotly dashboard, and support multilingual reports in English, French, and Spanish via Agentic AI (≥97% accuracy).

github

Biometric Aware Fraud Risk Dashboard with Agentic AI Avatar:

Description:

This application creates biometric signals, behavioral analytics, and an Agentic AI Avatar power this real-time fraud detection dashboard. It delivers sub-2-second latency, boosts precision by 42%, and cuts manual review time by 60%. More than detection—it guides compliance teams through risk with avatar-led clarity.

github

Arctic Sentinel: AI Native ISR Dashboard:

Description:

The application uses an ISR dashboard provides secure telemetry with response times under one second, achieves over 95% anomaly detection accuracy, and reduces workload by 40% through machine learning forecasts and three-dimensional overlays—enabling more informed and timely decision-making.

github

Arctic BlueSense: AI Powered Ocean Monitoring:

Description:

An AI system for the Canadian Arctic delivers sub second preprocessing, >95% anomaly detection accuracy, and 100% audit trace coverage secured with RS256 encryption. It monitors vessel activity and climate driven water changes while an agentic assistant provides real time alerts and explanations.

github

System Stability and Performance Analysis:

Description:

The system uses AWS Lambda and Agentic AI to diagnose issues in real time, secure logins with RS256, and restore sessions automatically. It delivers 98%+ crash free sessions, a 15% rise in login success, faster recoveries, 40% automated fixes, and a 30% drop in support tickets—boosting reliability, security, and user trust.

github

Sentinel Threat Wall:

Description:

Sentinel ThreatWall combines a fast C++ firewall with AI-based anomaly detection, using real-time inspection and graph analysis to reveal hidden traffic patterns. Its agentic layer interprets anomalies, correlates signals, and suggests adaptive defenses, employing RS256-signed telemetry and rule distribution for security. The platform offers sub-5 ms processing, over 92% classification accuracy, 95% clustering resolution, and maintains throughput above 1 Gbps.

github

Autonomous Security Orchestration Layer:

Description:

This application is a self evolving defense model that analyzes behavioral telemetry and system diagnostics to generate real time, context aware countermeasures. It deploys RS256 signed “digital antibodies” within seconds, supported by a browser based 3D operational dashboard built with JavaScript, HTML, and a dedicated 3D frontend layer for live system visualization. Early results show a 47% reduction in dwell time, 39% faster containment, 28% higher behavioral detection accuracy, and 31% more policy consistent responses.

github

Continuous learning:

I am a lifelong learner with a passion for staying ahead of the curve. I have deepened my expertise in Cybersecurity through Coursera and Data Scientist: Natural Language Processing as part of Codecademy’s online program. My curiosity fuels a continuous pursuit of knowledge, especially in rapidly evolving fields such as Agentic AI, large language models (LLMs), Generative AI, CLoud, Software and Cybersecurity. I strive not only to understand cutting-edge technology, but to apply it meaningfully and responsibly.

Work Experience:

Data Scientist (Freelance - UpWork)

February 2025 - Present:

• Delivered AI, LLM, and Python solutions across multiple client projects, automating workflows and improving data processes.

• Built and optimized Pandas/NumPy pipelines, reducing project timelines by 20% through efficient data modeling and cleaning.

• Developed JavaScript web apps for client analytics, improving UI responsiveness and cross device performance.

• Improved C++ data processing components, boosting efficiency by 20% and reducing runtime errors by 30%

Technical Expertise:

I have hands-on experience with various programming languages and technologies:

I thrive on tackling complex challenges, driving innovation through collaboration, and pushing the limits of technology to create meaningful impact.

Anthropic Courses:

AWS Cloud:

Dev.to (Articles and Projects):

dev.to

Education:

Volunteer:

OWASP Ottawa Chapter (Cybersecurity), Volunteer Staff: February 2025 - Present:

Attend IT Industry Network Event:

Video Content by Ben Nguyen on YouTube (Cybersecurity, AI, and Software):

YouTube

YouTube

Youtube

Youtube

Hackathon:

Hackathon 2025 (AI Tinkeres Ottawa - Build to Convert):

Our team has launched a website offering affordable meal plan subscriptions, supported by an Agentic AI Avatar that allows users to ask questions about the price of the meals in English or French.

Hackathon 2024 (NASA Space Apps Challenge_Hackalthon_2024.pdf):

I led a team to assess flood impacts across Canada by utilising ESRI mapping for geospatial visualisation, a modular backend for data processing, and a web-based frontend designed for stakeholder engagement.

Contact Information:

LinkedIn

Email: nguyenben85@gmail.com