Alexander Peroulas

Alex

Second-year CS student at UC Santa Barbara. I write software part time at LMI, where I built a tool that converts structured engineering models into runnable simulations — still in daily use by the engineering team. Strongest in C++, ship mostly in Python.

Download Resume

About Me

Second-year CS student at UC Santa Barbara. C++ is my strongest language, though most of what I’ve shipped is in Python. I’ve built LLM tooling and a custom MCP server, analytics dashboards over a few million rows, and the interfaces that sit on top of them. In between I work on things for myself: a chess engine, a movie search algorithm, a stat tracker.

I write software part time at LMI, where I was the sole developer on a tool that converts structured engineering models into runnable simulations, scoped from a one-line directive. It removes 2-3 hours of manual engineering work a day and is still in daily use by the engineering team.

While building it I ran into a smaller problem nobody had assigned me. Engineers were responsible for reviewing models they couldn’t open, because the software licenses didn’t reach everyone whose sign-off the work needed. So I built a viewer that reconstructs the diagrams from parsed model data and needs no license.

I also work with a team on Cadence, an AI interview coach, where I built the TwelveLabs video analysis integration and a Gemini pipeline that scores verbal responses.

Currently looking for Summer 2027 software engineering internships.

Experience

LMI logo

June 2026 – Present · Colorado Springs, CO

Space Systems Software Developer (Intern → Part Time)

LMI

2–3 hrs saved per day 500+ mappings 5,000+ lines, solo

  • Sole developer of a full-stack SysML-to-simulation conversion tool (Python, Flask, 5,000+ lines), scoped with the engineering team from a one-line directive: extracts models through a custom Cameo report, reconstructs their structure as JSON, and outputs runnable mission simulation files with version-to-version change logs. Still in daily use by the engineering team, removing 2-3 hours per day of manual work.
  • Resolved parameter mappings between schemas with no structural or naming correspondence, matching reconstructed model structure against both field names and their protobuf and JSON schema implementations rather than on labels alone.
  • Served a locally hosted Gemma model (Ollama) through a custom Python MCP server, decomposing each comparison to fit a small context window and ranking candidates by confidence across 500+ parameter mappings in testing.
  • Created a standalone model visualization tool that reconstructs BDD, activity, and state machine diagrams from parsed model data, giving engineers without a commercial modeling license access to models they were responsible for reviewing.
  • Developed a full-stack mission analytics dashboard using MongoDB, PyMongo, and Plotly, supporting interactive exploration across millions of data points with dynamically generated plots.

Fall 2026 – Present

Undergraduate Researcher — Early Research Scholars Program (ERSP)

UC Santa Barbara

Selected for competitive undergraduate research program; project assignment pending.

NTHS logo

2023 – 2025

President — National Technical Honor Society (NTHS)

Led chapter operations, coordinated events, and supported technical and academic initiatives, increasing chapter membership by 15%.

2023 – 2025

Team Captain — CyberPatriot

Led a four-person team hardening Windows and Linux environments (firewall configuration, vulnerability remediation, access control), earning Platinum Tier (top 30% nationally) across consecutive seasons.

Academy School District 20 logo

Fall 2021 – Spring 2025

Information Technology Specialist

Supported school IT operations, assisted staff and students, and managed technology and library systems.

Education

University of California, Santa Barbara

B.S. Computer Science, College of Engineering

Second year · Expected 2028

GPA: 3.86

Relevant Coursework

  • CMPSC 24 – Data Structures
  • CMPSC 32 – OOP, Unix Systems
  • CMPSC 40 – Discrete Math
  • CMPSC 130A – Algorithms & Data
  • CMPSC 138 – Automata & Formal Languages
  • CMPSC 156 – Software Engineering
  • PSTAT 120A – Probability & Statistics
  • Linear Algebra
  • Differential Equations

Skills & Certifications

Languages

  • C++
  • Python
  • Java
  • C
  • C#
  • SQL

Frameworks & Tools

  • Flask
  • Plotly
  • PyMongo
  • Git & GitHub
  • Jira
  • Figma

Technologies

  • Model Context Protocol (MCP)
  • Prompt Engineering
  • MongoDB
  • JSON

Certifications

  • CompTIA Cloud Essentials+
  • CompTIA IT Fundamentals+ (ITF+)
  • Java Certification — Certiport
  • Databases Certification — Certiport

Projects

Cadence — AI Interview Coach

Jan – Feb 2026 · Team of 4 (SB Hacks) · TypeScript, Gemini API, TwelveLabs API, Supabase

  • Integrated the TwelveLabs Marengo video analysis API to evaluate nonverbal interview performance, tracking eye contact consistency, hand gesture frequency, posture shifts, and speech confidence from browser-recorded video in real time, with all processing client-side so recordings never left the browser.
  • Engineered a prompt pipeline using Gemini 2.5 Flash to assess verbal response quality, flagging filler words, awkward phrasing, and pacing issues while scoring technical accuracy per question.
  • Contributed to a post-interview feedback dashboard with Supabase-backed session history, letting users track scores across practice sessions.
View Project GitHub

Contact Me

Reach out via LinkedIn or email.

Email: [email protected]

LinkedIn GitHub