AI-Powered Academic Document Generator

Create professional reports, presentations, conference papers, academic theses, and AI images in seconds. Powered by advanced AI with automatic formatting, citations, and free image generation.

Comment

Inspiration

As engineering and research students, we realized that valuable time is often lost not on research, but on formatting. Wrestling with LaTeX configurations, fixing citation styles, creating Gantt charts, and ensuring compliance with strict standards (like BET or IEEE) takes hours. We wanted to build a tool that handles the tedious administrative side of academia—formatting, structuring, and citing—so students and researchers can focus entirely on innovation and content.

What it does

AcademicGen is an all-in-one platform that instantly generates professional academic documents and resources. It goes beyond simple text generation by enforcing strict academic standards.

  • Technical & Custom Reports: Generates BET-standard reports complete with Gantt charts, budget tables, and LaTeX equations.
  • Conference Papers: Creates IEEE-formatted papers with proper two-column layouts, Roman numeral headings, and abstract/keyword sections.
  • Theses & Dissertations: Drafts full chapters with methodology, results, and Harvard citations.
  • Smart Presentations: Generates PowerPoint slides following the 6x6 rule, including speaker notes and AI Text-to-Speech (TTS) coaching to help you practice your delivery.
  • Visuals: Includes an AI Image Generator (via Pollinations.ai) and auto-fetches relevant stock photos (via Pixabay).
  • Reference Library: Manages citations with smart imports from URLs, DOIs, and BibTeX.

How we built it

The core of AcademicGen is powered by Google Gemini 2.5 Flash, which allows for rapid, high-context generation of complex technical content.

  • Backend: We utilized Gemini 2.5 Flash for its superior reasoning in structuring academic arguments and handling LaTeX syntax.
  • Image Generation: Integrated Pollinations.ai for text-to-image creation and the Pixabay API for fetching relevant stock assets for presentations.
  • Formatting Engine: We built custom parsers to ensure output strictly adheres to IEEE and BET formatting standards.

Challenges we ran into

  • Citation Accuracy: Ensuring the AI didn't just hallucinate sources but correctly formatted references (DOI, BibTeX) was difficult. We had to implement a strict verification layer for the Reference Library.
  • Complex Formatting: Teaching the model to consistently output valid LaTeX for equations and proper two-column layouts for IEEE papers required extensive prompt engineering and fine-tuning.
  • Integration: Seamlessly stitching together text generation, image creation, and TTS audio into a single cohesive user experience.

Accomplishments that we're proud of

  • Successfully implementing 7 different document types and over 66+ topic templates.
  • The TTS Coaching feature: It doesn't just write your slides; it literally teaches you how to present them.
  • Achieving a level of quality where the output requires minimal editing to meet BET and IEEE standards.
  • Building a tool that creates Gantt charts and Budget tables automatically—a huge time-saver for engineering students.

What we learned

We learned the immense potential of Gemini 2.5 Flash in handling structured, rule-based creative tasks. We also discovered that "academic tone" is subjective, which led us to implement style controls. Most importantly, we learned that students don't just want text; they want finished products—formatted, cited, and ready to submit.

What's next for AI-Powered Academic Document Generator

  • LMS Integration: Plugging directly into Canvas or Blackboard for seamless assignment submission.
  • Collaborative Editing: Allowing multiple users to work on a generated thesis or report in real-time.
  • More Citation Styles: Expanding beyond IEEE and Harvard to include APA 7, MLA, and Chicago styles.
  • Mobile App: A companion app for reviewing generated documents and practicing presentations on the go.

Built With

Share this project:

Updates