Aleksander Kolasinski

Software Engineer. Always learning.

I build software across the full stack. I've shipped production APIs, data pipelines, and client-facing sites for early-stage startups. I also freelance on the side.

Open to full-time software engineering roles.

Work Experience

Software Engineer, FreelanceLa Brea Law, APC

June 2026 – Present

Built marketing site from Figma mockups using Next.js, React, and Tailwind CSS.
Implemented contact form with server-side email delivery and client/server-side validation.
Improved Lighthouse score from 65 to 92; configured SEO metadata, Open Graph tags, and production security headers on Vercel.

Next.jsReactTailwind CSSResendZodVercel

Full Stack Software Engineering InternAtrium Energy

Feb 2026 – Apr 2026

Built production REST API endpoints with FastAPI and Python, integrating MongoDB to automate customer onboarding workflows on AWS EC2.
Implemented JWT/JWKS authentication securing backend CRM services across a microservice architecture.
Designed a time-series pipeline connecting MongoDB and TimescaleDB to serve property analytics to a React frontend.

FastAPIPythonMongoDBTimescaleDBReactAWS EC2JWT

Frontend Developer InternFamous Title, LLC

Nov 2023 – Jun 2024

Delivered a client-facing website for a medical practice with a responsive UI using HTML, CSS, and JavaScript.
Built a contact form that automatically emailed patient inquiries to the doctor.

HTMLCSSJavaScript

Projects

Habit Tracking Web AppGitHub

Built a full-stack web app using Node.js, Express.js, and PostgreSQL to manage daily habit tracking with persistent data storage.
Designed a RESTful API with CRUD operations and a relational database schema.
Created a responsive React/TypeScript frontend with modular architecture and asynchronous data fetching.

Node.jsExpress.jsPostgreSQLTypeScriptReact

BiteBook Food Journal

Led a team of five to build a cross-platform food journal in Flutter with photo uploads, notes, and daily tracking.
Implemented Firebase Authentication and Storage for secure image handling and real-time sync.
Added sharing via dynamic links and QR codes, allowing users to duplicate entries into their own journals.

FlutterFirebase

Smoking Image Classification

Built a deep learning pipeline using custom CNNs and MobileNetV2 transfer learning to detect smoking behavior.
Improved performance through data augmentation, regularization, and ensemble modeling using TensorFlow and scikit-learn.
Achieved 88% test accuracy with high recall using precision-recall metrics and ROC analysis.

PythonTensorFlowMobileNetV2scikit-learnNumPypandas