The Campbell's Recipe API project is designed to facilitate personalized recipe suggestions for users through a conversational interface on a website. The system leverages REST API calls to manage the dialogue, apply guardrails, recognize user preferences, and ultimately deliver curated recipe options based on the interaction.

Technologies: OpenAI, NodeJS, AWS, DynamoDB, Python

This project aims to augment the capabilities of ChatGPT by integrating advanced document management functionality and contextual enrichment features. The endeavor focuses on developing a robust wrapper around the core ChatGPT model that allows users to upload, manage, and reference documents within conversational flows. Additionally, it facilitates the creation of customizable personas and the incorporation of external context derived from documents to enhance the interaction quality and relevance.

Technologies: OpenAI, NodeJS, AWS, NextJS, DynamoDB, Python, PostgresSQL

This project is an interactive Alexa skill designed to engage users in convincing The Grinch to like Christmas, featuring personalized responses, easter eggs, and podcast discovery, targeting families and Grinch fans. The skill operates on a voice clone algorithm to maintain consistent character delivery, underpinned by a robust technical framework for real-time user engagement.

Technologies: OpenAI, NodeJS, AWS, ElevenLabs, DynamoDB

HeyEdu's AI Assistant facilitates user engagement in a new educational world by providing descriptions upon selection and confirmation, thereafter suggesting an initial learning path and dispatching tasks via email, with the Assistant available for support as needed. The system emphasizes practical application of knowledge, employing gamification elements to motivate and guide users towards competency building and achieving their goals.

Technologies: OpenAI, NodeJS, AWS, DynamoDB

RescueBot is an mobile application designed to provide real-time guidance in emergency situations, offering straightforward, precise first aid instructions via an interactive voice interface that tailors advice based on user responses to critical questions. It aims to bridge the critical gap between accident occurrence and emergency services arrival, enhancing survival chances through immediate, informed action, and supports both Polish and English to ensure accessibility across diverse user groups.

Technologies: OpenAI, NodeJS, AWS, Flutter, DynamoDB

HearMe.ai is a text-to-speech conversion platform designed to turn written content into natural-sounding audio, enabling publishers to reach wider audiences through major podcast platforms and direct website integration. It offers plans with different monthly character limits and requires no technical skills for use, catering to everyone from individual bloggers to large-scale content creators seeking to make their content more accessible and engaging.

Technologies: OpenAI, NodeJS, AWS,  AWS SES, AWS SQS, AWS CloudFront, Wordpress, GoogleSpeech, PostgresSQL

The project delivers a dynamic reporting system for road data to a government road agency, facilitating real-time traffic monitoring, incident analysis, and maintenance tracking to improve roadway management and safety. It includes interactive dashboards, data visualization, and automated reporting functionalities.

Technologies: .NET, ReactJS, Docker, Jenkins

Road managers and owners can operate more efficiently if they monitor and analyse traffic. Verified data, collected during traffic measurements, is essential for reliable analysis. Organizing a measurement campaign alone is a huge challenge due to the large amount of data collected from hundreds of measurement points from different contractors. Our Vera platform, dedicated to road managers and owners, allows efficient management of measurement campaigns and unification of the results obtained.

Technologies: Java, Hibernate, Jenkins