Projects Welcome to the Projects section of StarLit(b)Log! This is where I document my ongoing explorations, experiments, and deep dives into topics that interest me. Some of these are purely research-driven, some are hands-on technical builds, and others are just me following a curiosity until it leads somewhere useful (or at least interesting!). Each project has a starting point, but not necessarily an “end”—things evolve, new ideas emerge, and sometimes I revisit old projects from a fresh angle. Below are the current projects I’m working on or have explored in depth. Queryable Web The web is an enormous resource of information, but most of it remains unstructured and difficult to query in a meaningful way. The Queryable Web project is my deep dive into making web data more accessible and structured, inspired by research on web data extraction and retrieval systems like the Doctor Web Engine. The challenge isn’t just retrieving information; it’s about handling dynamic content, dealing with messy data, and structuring responses in ways that make the web behave more like a queryable database. In this series, I explore different approaches—leveraging web scraping, semantic web technologies, and methods like DOM parsing, regular expressions, and structured query languages. While solutions like RDF and semantic web standards offer one path, widespread adoption has been slow, meaning alternative strategies are needed. My goal is to experiment with query-based data extraction tools, tackle real-world challenges in web information retrieval, and explore how we can bridge the gap between raw web content and structured, searchable data. Movie Transcript GPT Model This is a pro bono research project I’m working on in collaboration with Birkbeck, University of London, involving both the School of Mathematical and Computing Sciences and the School of Arts. The idea behind this project is to explore how GPT models can be fine-tuned using movie transcripts, studying how AI learns from cinematic storytelling, dialogue, and character interactions. By bridging computational techniques with artistic narratives, we’re investigating how well AI can generate realistic dialogue, understand emotional context, and mimic distinct storytelling styles . This project isn’t just about training a model—it’s about exploring the creative and technical limits of AI in film and narrative structures. Through this series, I’ll be documenting our approach, findings, and how AI can contribute to the world of storytelling in meaningful ways. Research Review This is where I geek out over research—both my own and from the broader academic community. As part of my doctoral work, I’ve explored topics around AI, trust in data, and information retrieval, and this space is where I share findings from my published papers as well as my thoughts on research that inspires me. Beyond my own work, I use this series to review and break down complex academic papers in a way that’s easier to digest. Whether it’s cutting-edge AI models, search algorithms, or ethical debates around technology, I explore topics that I believe are worth discussing. If you’re into deep technical reading mixed with hands-on applications, you’ll feel right at home here. Home lab There’s something satisfying about running your own infrastructure, fine-tuning it to your exact needs, and experimenting with self-hosted systems. My Home lab series is all about building and managing a personal tech playground—tinkering with server setups, optimizing networks, and running cool projects that I otherwise wouldn’t get to experiment with in a cloud-based world. I’ll be documenting my journey in configuring hardware, setting up automations, experimenting with self-hosted software (e.g. this (b)log), and generally pushing my home server setup to do more. Whether you’re deep into home lab culture or just curious about running your own server, this is my little corner for tech experiments that live outside the cloud.