Hristijan – Python, React, Node.js

An AI teammate for property teams, helping streamline leasing and residential real estate operations. Based in London, the company operates in the leasing residential real estate space with a team of 11-50 employees.

• Worked as a full-stack engineer with end-to-end feature ownership across the entire stack.
• Designed and shipped core AI-powered features including fully automated pre-vetting and maintenance agentic workflows.
• Built a custom embeddable chatbot with white-label branding and custom AI tools for fetching live platform context.
• Built a viewing booking engine with intelligent slot suggestions and connected it to an AI tool enabling the LLM to autonomously offer and book slots in conversation.
• Contributed to prompt engineering.
• Worked with RAG including document chunking to surface relevant content within AI responses.
• Delivered third-party integrations spanning Google Calendar, Microsoft Calendar, Calendly, Twilio, WhatsApp, Res:harmonics, Arthur, RightPlace, Coho, PipeDrive, Spike, and GlueDog.
• Built a message templates feature with dynamic placeholders.
• Used Inngest for task queuing and email notifications.
• Handled complex UI forms.
• Contributed to UI redesign efforts across the product.

An IT services and consulting company focused on software, data, and partnerships, aiming to deliver long-lasting solutions by solving hard problems rather than simply selling services.

• Joined the front-end team of a logistics platform and contributed to active feature development while porting functionality from a legacy desktop application into a modern React web interface.
• Designed and implemented a translation parser to centralize localization handling across the application.
• Worked extensively with TypeScript, Redux, Styled Components, and custom UI component development.
• Built a multi-step modal with full validation logic and a suite of reusable form components.
• Worked on the front-end of an internal platform used to test proprietary network equipment.
• Led the migration from Vue 2 to Vue 3.
• Replaced the legacy Webpack configuration with Vite.
• Moved the global state layer from Vuex to Pinia.
• Optimized data loading and rendering across the UI, reducing wait times by over 70%.
• Introduced reusable component libraries built to a design system.
• Architected data fetching abstractions with custom API transformers to keep the UI layer decoupled from raw backend responses.
• Mentored four interns across two consecutive summer programmes, guiding them through MongoDB, TypeScript, Express, NestJS, Python, and basic shell scripting.
• Independently built a Pipedrive–Apollo integration tool in TypeScript, packaged as a standalone executable binary designed to be scheduled via cron or triggered manually.
• Contributed to the GrabIT company website built on a Laravel-based CMS.
• Delivered new features including a blog, carried out design improvements, and performed targeted SEO optimizations.

An IT services and consulting company focused on software, data, and partnerships, aiming to deliver long-lasting solutions by solving hard problems rather than simply selling services.

• Explored and implemented recommendation systems, training a model using TensorFlow.
• Gained hands-on exposure to deep learning frameworks, applying both collaborative and content-based filtering techniques.

An IT services and consulting company focused on software, data, and partnerships, aiming to deliver long-lasting solutions by solving hard problems rather than simply selling services.

• Gained hands-on experience in the end-to-end machine learning workflow, starting from exploratory data analysis and progressing through data cleaning, feature preparation, and model training across multiple datasets.
• Designed and developed a full-stack ML pipeline web application using Flask, HTML, CSS, and JavaScript that allowed loading raw CSV data, performing data transformations, training predictive models, and evaluating performance through visual metrics.
• Built a churn prediction model for a telecom operator, leveraging both a manual workflow in code and the full-stack app to train, evaluate, and run predictions.