Erik Sherman
Software Engineer & Entrepreneur
note: yes, I did make this website myself! using nextjs + typescript
I left university after two years to focus on my thriving software development agency, having won multiple six-figure contracts that demanded my full attention. As lead developer and project manager, I built a pipeline to recruit and evaluate overseas talent and assemble teams to complete projects in React + TypeScript. I later founded cleat.ai, a software platform tailored to government contractors enabling them to find, bid, and win government contracts using AI.
Work Experience
Cleat.ai
Founder
- Developed an AI-powered platform for government contractors
- Built tools to streamline contract discovery, bidding, and acquisition processes
- Leveraged cutting-edge AI technology to optimize contract matching and success rates
Dulles Glass and Mirror
Lead Developer
- Led complete rebuild of legacy e-commerce system with a focus on performance and scalability
- Designed and built a shower door customizer with extensive glass customization options
- Measured detailed usage statistics and performed performance optimizations to increase sales
- Collaborated with client's existing team to integrate with ERP system that manages their factory
MeTime Healing
Lead Developer
- Developed a wellness platform for scheduling and managing therapy sessions
- Implemented real-time booking and calendar synchronization
- Built secure payment processing integration
Education
University of Maryland
College Park, MD
Computer Science Major (Left to pursue software development agency dashrd.com)
Relevant Coursework
Activities
- Terps Racing Baja - Designed and manufactured suspension component jigs
- Completed specialized CNC mill training
- Used SolidWorks, Fusion 360 (CAD and CAM)
Wootton High School
Rockville, MD
Achievements
- Excellence in Computer Science Senior Superlative
- GPA: 3.87 / WGPA: 4.58
- AP Scholar with Honor
- SAT Math: 770, Physics: 790
Relevant Coursework
AI & Machine Learning
Explored deep neural networks at the model level, working directly with PyTorch to build and train custom architectures. Leveraged GCP GPUs for training and experimented with techniques such as transfer learning and model fine-tuning. Focused on practical implementation of state-of-the-art deep learning methods.
Courses
- Deep Learning for Coders (Jeremy Howard) - fast.ai
- Deep Learning Specialization (Andrew Ng) - Coursera
Projects
Face Detection with Transfer Learning
Built a PyTorch-based facial detection system using transfer learning from pre-trained ImageNet models. The system fine-tunes the model to predict bounding boxes around faces, demonstrating practical application of transfer learning techniques.