Resume

Download my resume to learn more about my experience and qualifications.

VALENTINA TORRES DA SILVA

valentinatsilva@proton.me | LinkedIn | GitHub | Portfolio | New York City, NY

EDUCATION

Fairleigh Dickinson University

Teaneck, New Jersey

Bachelor of Science in Computer Science | Anticipated May 2026

TECHNICAL STACK

Programming Languages

Python, R, Java, JavaScript, Bash

Frameworks & Libraries

Scikit-Learn, TensorFlow, PyTorch, Next.js, TailwindCSS, FastAPI, Pandas, NumPy, Matplotlib, Seaborn

Databases

SQL, PostgreSQL (via Supabase)

Cloud Platforms

AWS (SageMaker, Lambda, API)

Data Skills

NLP, Supervised Machine Learning (Regression & Classification), Retrieval-Augmented Generation (RAG)

DevOps & CI/CD

Linux, VS Code, Cursor, GitHub Copilot

PROJECTS

Long Run | Battle of the Brains "Best Technology Solution" Winner | Team Leader

October 2024

  • Secured 3rd place out of 30+ teams by delivering a high-performance React.js + Next.js platform in 16 hours, optimizing client-side routing to reduce load times by 35% and improve user engagement
  • Directed a 3-member dev team to design and deploy a multi-page event platform with embedded video, interactive scheduling, and eco-materials showcase, meeting competition criteria under tight deadlines
  • Presented technical and design rationale to the president of KPMG, reinforcing project credibility and solution viability

RAG-Enabled AI-Powered Legal Contract Data Extraction | FDU + RSG Media/ Rightsline

August 2024 - November 2024

  • Cut legal contract review time by 70% by developing a Retrieval-Augmented Generation (RAG) application that automated context-aware data extraction and accelerated content generation
  • Built a FastAPI back end with PostgreSQL (via Supabase) to deliver near-real-time query results, reducing retrieval latency by 65% and improving data accessibility for stakeholders
  • Designed a responsive, TailwindCSS-styled Next.js front end, boosting usability and achieving 100% higher satisfaction ratings during stakeholder demos
  • Enhanced AI output accuracy by 90% through OpenAI API integration with prompt engineering, resolving critical gaps in contract processing workflows

Detection of AI-Generated Social Bots on Twitter

April 2025 - May 2025

  • Designed an end-to-end pipeline across classical ML, deep learning, and clustering: preprocessing (English-tweet filtering ≥70%, normalization), feature engineering (URL/hashtag/mention counts, profile signals, account age, friends-to-followers), scaling, and consistent splits
  • Benchmarked KNN (0.86), Logistic Regression (0.87), SVM (0.87), Random Forest (0.86), Bi-LSTM (0.87), and PCA+KMeans clustering (0.88), confirming DL outperformed classical ML on precision/recall/F1 while clustering delivered the top accuracy without labels

News Headline Classification with AWS SageMaker

August 2025

  • Fine-tuned DistilBERT on headline dataset using PyTorch + Hugging Face, achieving 96% validation accuracy (loss 0.118)
  • Deployed a real-time inference API with SageMaker, Lambda, and API Gateway, delivering predictions in ~1.5s latency
  • Integrated CloudWatch monitoring and S3 model storage to ensure production-ready reliability and cost control

WORK EXPERIENCE

Fairleigh Dickinson University

Robotics Lab Assistant | August 2024 - Present

  • Improved NAO humanoid AI robot task accuracy by 40% by configuring custom Python scripts for multi-step automation workflows
  • Resolved 12 technical challenges by designing creative code-based solutions, accelerating lab experiment completion and ensuring operational readiness for research demonstrations
  • Researched and implemented emerging AI and robotics techniques to optimize task execution speed and robot adaptability

Fairleigh Dickinson University

Computer Lab Assistant | August 2023 - Present

  • Reduced support request resolution time by 50% by diagnosing and resolving software, hardware, and login issues for 50+ faculty and students weekly
  • Maintained and updated lab equipment, ensuring 100% availability and operational efficiency during peak usage periods
  • Delivered technical assistance and user guidance that improved end-user satisfaction scores by 20%

AWARDS

eBay Innovation Scholarship – $5,000

March 2024

  • One of only 4 students selected out of hundreds for an essay demonstrating the measurable community impact of a completed technology project

GHSCSE Directors Discretionary Fund Scholarship

Spring 2025

  • Recognized for exceptional academic performance and contribution to the computing community

NCAA Division-1 Student Athlete Awards

NEC All-Rookie Team 2022, Second Team All-Conference 2023

  • Honored for consistent athletic excellence while balancing full academic load, demonstrating resilience, leadership, and discipline transferable to technical roles