-
-
UI Page
-
Evidence Tab
-
Publication Strategy Tab
-
Insights Tab
Inspiration Thesis Guard was inspired by my lab workflow. I regularly spend time refining research ideas, finding papers, checking datasets, and deciding whether a topic is strong enough for a thesis or paper. I built it with my own work and my colleagues in mind.
What it does Thesis Guard helps turn a research topic into a structured plan. It retrieves papers and datasets, supports manual evidence input, lets users pin important sources, suggests publication targets, generates citation-aware writing support, and saves runs for iteration.
How I built it I built it with a FastAPI backend and a React frontend. The backend handles retrieval, structuring, planning, and local storage, while the frontend provides an interactive workspace for overview, evidence, publishing, citations, insights, and collaboration. I also integrated Amazon Nova, with a mock mode for demos and development.
Challenges I ran into The biggest challenge was relevance. Search results can seem related without actually being useful, so I had to improve filtering and interpretation. Another challenge was simplifying the interface and removing features that added noise instead of value.
Accomplishments that I'm proud of I am proud that the project became a practical research workflow tool rather than just a text generator. It supports evidence curation, dataset discovery, publication planning, citation help, and iterative saved runs in one place.
What I learned I learned that retrieval alone is not enough. The real value comes from interpretation, prioritization, and presenting results in a way that researchers can actually use.
What's next for Thesis_Guard Next, I want to make the system more domain-aware, improve collaboration features, and strengthen export and drafting support so it becomes more useful for real research teams.
Log in or sign up for Devpost to join the conversation.