I am Patryk Kuchta, an aspiring
Artificial Intelligence Scientist.
What I'm Up To
I spend much of my time designing and evaluating AI systems, or systems where AI plays a central role. My work involves the full lifecycle of AI implementation, from initial concept through to production deployment.
I have delivered training and adoption materials that have significantly improved AI usage within Softwire, helping teams understand both the capabilities and limitations of AI technologies. I've also created autonomous AI pipelines that have reduced developer burden by automating tasks that were previously time-consuming manual processes.
Working with diverse clients (from the highly skeptical to the enthusiastically optimistic), I've developed expertise in communicating AI benefits while respecting organizational constraints. This involves not just technical implementation, but also the human and ethical dimensions of AI adoption.
A particular area of passion for me is Ethical AI and Sustainable AI. I've spent considerable time advocating for sustainability considerations in AI systems, conducting environmental assessments for embedded AI, and building adoption systems that leverage local open-source AI instead of relying on large providers whose ethical standards can be questionable.
Education
- Graduated with Distinction
- Delivered multiple complex AI projects
- Conducted in-depth research projects in the NLP field
- Graduated with First Class Honours
- Recipient of the Westfield Trust Academic Prize
- Graduated with overall average at 90%
Work Experience
Growing up in Warsaw, I moved to London at 18. Since then, my career and education has taken me through Edinburgh and Manchester, shaping my journey across the UK.
Manchester
at Softwire
in Manchester, United Kingdom.
Since November 2025
at Softwire
in Manchester, United Kingdom.
November 2024 - November 2025
Edinburgh
at DataAnnotation Tech
in New York, United States (Remote).
Since January 2024
Greater London
at Queen Mary University
in London, United Kingdom.
September 2021 - April 2024
at Softwire
in London, United Kingdom.
June 2023 - August 2023
Warsaw
at McDonald's
in Warsaw, Poland.
July 2018 - August 2018
Certificates and Awards
Modules Taken
Human languages
Programming Languages
Frameworks/Libraries
Miscellaneous
Projects
Efficient Agents with Hot-Swappable Adapters
Developed a modular multi-agent system using parameter-efficient fine-tuning to enhance tool use in small language models. Implemented hard routing with expert adapters, achieving performance surpassing larger closed-source models. Demonstrated skills in distributed training, model optimization, and API integration.
AI Property Valuation with Novel Normalization
Built a deep learning model combining CNNs and structured data for property valuation, achieving R² of 0.7287. Introduced novel geo-temporal normalization for market dynamics. Applied transfer learning, advanced regularization, and multi-modal data fusion including transport and POI maps. Led data collection and preprocessing.
Research into Multi-Agent LLM Tool Use Strategies
Research proposal investigating parameter-efficient fine-tuning strategies for multi-agent LLM systems. Designed novel task division approaches to enhance tool use in small language models. Demonstrated research design, literature synthesis, and experimental methodology planning.
Research Review: Neural Low-Resource Language Translation
Conducted comprehensive research review of neural machine translation techniques for low-resource languages. Critically evaluated state-of-the-art approaches, identifying key challenges and opportunities. Demonstrated research synthesis, technical writing, and analytical evaluation skills. Received excellent marks across all criteria.
Analysis of PSO, GA and GP Algorithms
Implemented and analyzed three bio-inspired optimization algorithms: Particle Swarm Optimization, Genetic Algorithms, and Genetic Programming. Conducted comparative performance analysis across different problem domains. Demonstrated algorithm implementation, experimental design, and data analysis skills.
Learnopedia: Collaborative Learning Platform
Developed full-stack collaborative learning platform with democratic course creation system. Implemented user authentication, content management, and real-time collaboration features. Conducted competitive analysis and user testing. Demonstrated React, TypeScript, Express.js, and database design skills.
Deep Learning Image Classification (CIFAR-10)
Built CNN-based image classifier achieving 95.5% accuracy on CIFAR-10 dataset. Applied deep learning techniques including data augmentation, regularization, and hyperparameter tuning. Demonstrated model training, evaluation, and optimization skills. Scored 100% on coursework assignment.
Complete Prototype of a Cryptocurrency Wallet
Led team of 6 to build cryptocurrency wallet prototype from domain analysis to deployment. Managed project timeline, task allocation, and code integration. Developed RESTful API and full-stack application using React and Express. Won best project award for technical execution and teamwork.
PhotoCa.st: Weather App for Photographers
Developed photographer-focused weather application with minimalistic UI/UX design. Integrated external weather APIs and implemented responsive interface. Led team project, managing development workflow and stakeholder requirements. Demonstrated API integration, UI design, and project management skills.
Psychotherapist Portfolio Website
Built fully responsive portfolio website for psychotherapy practice using React and TypeScript. Collaborated with client to design calm, professional interface. Implemented accessibility standards and cross-device compatibility. Demonstrated client communication, responsive design, and modern web development skills.
Architect Portfolio Website
Developed portfolio website for architecture firm, translating design sketches into functional code. Implemented interactive project gallery with JavaScript and responsive layouts using Bootstrap. Demonstrated design-to-code translation, front-end development, and mobile-first design principles.