Agentic workflow design
Designing multi-agent systems that translate ERP change requests into test cases and executable PeopleSoft Test Framework scripts.
Computer Science at the University of South Carolina
Hi, I'm Vito Spatafora.
I'm a computer science student working across AI consulting, machine learning research, and enterprise automation. I like building practical systems that connect technical ideas to work people actually need done.
8
interns coordinated as IBM Columbia capstone lead
30h
monthly manual work reduced with PowerShell automation
Top 4
Broadview intern recognition among a 40-person cohort
2027
expected B.S. Computer Science graduation
Profile
Building AI systems that make complex work easier to execute.
I am a Computer Science student at the University of South Carolina with experience applying AI, machine learning, and software engineering to business and research problems. My work sits at the intersection of technical implementation and practical impact: understanding the workflow, designing the system, and turning the idea into something people can actually use.
At IBM, I develop AI-driven solutions for enterprise applications, including agentic workflows and automation tools. At USC, I contribute to machine learning research for real-time industrial anomaly detection using computer vision, sensor data, and deployable inference pipelines.
Real-time model deployment
Moving anomaly detection from offline batch processing toward streaming inference on constrained edge hardware.
Business-aware engineering
Collaborating with technical and business stakeholders to identify where AI improves speed, consistency, and scalability.
Skills
Tools I use to move from idea to working system.
Languages
Python
Java
C++
C#
JavaScript
SQL
PowerShell
HTML
AI & Machine Learning
Generative AI
Agentic Workflows
RAG
Vector Databases
Computer Vision
TensorFlow
Random Forests
Feature Engineering
Enterprise & Cloud
PeopleSoft
PTF Scripting
AWS Cloud
Automation
Knowledge Management
Workflow Analysis
Platforms & Practices
Git/GitHub
Linux
Windows
Jira
Excel
Technical Reporting
Experience
Recent roles across AI consulting, research, operations, and leadership.
Associate Application Consultant Intern
- Utilize IBM AI tools and agentic workflows to conduct research, support consulting initiatives, and identify opportunities to improve internal processes and employee productivity.
- Lead development of an AI-driven automation solution for PeopleSoft Test Framework scripting, addressing a manual testing process that requires significant scripting effort.
- Design a multi-agent architecture that transforms PeopleSoft ERP change requests into natural language test cases, then into executable PTF scripts informed by documentation and best practices.
- Selected as Team Lead for IBM Columbia's intern capstone team, coordinating eight interns in a nationwide consulting case competition.
- Contributed to an AI-enabled knowledge management proposal for NASA using RAG, vector databases, and conversational AI to improve knowledge accessibility and retention.
Machine Learning Researcher | University Research Team
- Contribute to a machine learning system for anomaly detection in industrial welding processes using thermal imagery, visual imagery, and temporal sensor data.
- Support anomaly detection pipelines using feature engineering, threshold-based detection, and Random Forest models to identify irregular machine behavior and process defects.
- Led the transition from offline batch processing to a real-time inference environment for live anomaly detection from streaming camera data on edge devices.
- Develop and maintain Python data processing and ML workflows for deployment, testing, and performance evaluation.
- Conduct feature selection, preprocessing analysis, and dataset compatibility reviews across heterogeneous data sources.
VP of Health and Safety
- Developed and enforced safety protocols for events, including risk management strategies and compliance with organizational and university policies.
- Regularly met with university officials, alumni advisors, and national leadership to ensure chapter needs were addressed and members were supported.
- Spearheaded initiatives supporting the health, safety, and well-being of 100+ members while promoting accountability and care.
End Point Engineer Intern
- Maintained service records and prepared technical reports to document findings and preserve operational history.
- Supported senior engineers with troubleshooting, system calibration, installation, maintenance, and repair work.
- Imaged computers with a custom Windows operating system and automated inventory management with PowerShell, reducing manual data entry by approximately 30 hours monthly.
Member Solution Specialist
- Recognized as a top 4 intern among 40 interns for customer service and efficiency.
- Assisted members with personal finance workflows, including secure loan payments and fund transfers.
- Guided clients through loan and financial services application processes.
AI & Research
Selected work where models, automation, and business impact meet.
Enterprise AI
PeopleSoft Test Framework automation
Designing a multi-agent workflow that interprets ERP change requests, drafts natural language test cases, and generates executable PTF scripts using documentation-informed AI agents.
Agentic workflows
Prompt architecture
Enterprise QA
Machine Learning
Real-time industrial anomaly detection
Contributing to a research system for welding process anomaly detection across thermal imagery, visual imagery, and temporal sensor data, with a deployment path for edge devices.
Computer vision
Random Forests
Streaming inference
Consulting Case
NASA knowledge retention proposal
Led team analysis around information silos, document discovery, knowledge transfer, and institutional expertise loss during employee transitions.
RAG
Vector databases
Conversational AI
ML Efficiency
Energy-aware chess engine training
Researching whether curriculum-guided self-play can improve training efficiency versus traditional full-game self-play while tracking convergence, energy use, and engine strength.
TensorFlow
CNN evaluation
Self-play
Projects
Technical work showing algorithms, automation, and practical systems thinking.
Systems / Algorithms
Chess Engine Development
Designed and implemented chess software with a built-in engine capable of evaluating board positions and selecting strong moves.
- Implemented Negamax, alpha-beta pruning, transposition tables, Zobrist hashing, and magic bitboards.
- Engine performance ranks in the 70th percentile of human players based on Chess.com statistics.
Automation
Endpoint Inventory Automation
Built PowerShell automation for endpoint inventory workflows during Broadview Federal Credit Union's endpoint engineering internship.
- Reduced recurring manual data entry by approximately 30 hours per month.
- Supported workstation configuration, Jira request handling, and technical documentation.
Research / ML
Optimizing ML Training Energy Efficiency
Built TensorFlow CNN evaluation functions to compare curriculum-guided self-play and traditional full-game self-play inside a chess-engine training framework.
- Measured convergence speed, energy usage, and relative model performance under consistent conditions.
Education
Computer science foundation with added cloud and ML credentials.
University of South Carolina
Relevant coursework includes Machine Learning, Data Structures, Algorithm Design and Analysis, Software Engineering, Computer Networks, Vector Calculus, Discrete Mathematics, and Physics.
Study abroad: academic and cultural immersion in Florence, Italy during Spring 2026.
AWS Certified Cloud Practitioner
Machine Learning Crash Course
AWS Cloud Practitioner Essentials
Resume & Contact