Project Ereuna

AI-Powered Research Report Generator

Technologies
Python Streamlit Google Gemini OpenAI Anthropic Claude API BeautifulSoup NLTK ReportLab python-docx python-pptx fpdf2 Pillow pypdf requests googlesearch-python markdown textstat
Category Academic & Research Tools

Project Overview

Project Ereuna is an AI-powered research report generator that leverages multiple AI models to create comprehensive research reports. It supports multi-volume documents up to 200+ pages with hierarchical organization, advanced natural language processing, and automated citation management for academic rigor.

60%
Time Reduction
200+
Pages Supported
50+
Auto Citations

Key Features

🤖

Multi-Model AI Integration

Unified API supporting Google Gemini, OpenAI, and Anthropic Claude with fallback options and seamless model switching

📄

Multi-Format Export

Export to DOCX, and TXT with full formatting, tables, hyperlinks, and professional slide generation

📚

Intelligent Citation Management

Automated APA, MLA, Chicago, and Harvard citations with DOI links and comprehensive bibliography creation

🔍

Research-Aware Chatbot

Interactive Q&A based on generated content with web search fallback, conversation memory, and history tracking

📊

Quality Assurance

Readability analysis, grammar verification, table summarization, and content rigor checks for peer-review readiness

Hierarchical Generation

Checkpoint resume, iterative refinement, and modular processing for long-form document creation

Technical Implementation

$ Developed Streamlit-based application leveraging multiple AI models for comprehensive research report generation, including multi-volume documents up to 200+ pages with hierarchical organization

$ Integrated advanced natural language processing using NLTK for analyzing research topics, identifying key themes, concepts, and supporting auto-generated executive summaries via extractive/abstractive techniques

$ Implemented multi-stage report generation pipeline including literature review, methodology, findings, conclusions, and additional sections like executive summaries, table summarizations, and bibliographies

$ Created sophisticated citation management system automatically generating proper academic references in APA, MLA, Chicago, and Harvard formats with automated bibliography creation including DOI links

$ Built content synthesis engine combining information from multiple sources into coherent narratives with proper attribution, contextual building across sections, and web scraping for literature collection using BeautifulSoup

$ Engineered customizable templates supporting various research paper formats (e.g., scientific, business, literary) with automated section structuring, JSON-based templates, and prompt management for different tasks

$ Implemented iterative refinement workflow allowing users to provide feedback, regenerate specific sections, and utilize checkpoint resume for long-running tasks to save/recover progress

$ Created export functionality supporting multiple formats including DOCX (with full formatting, tables, hyperlinks), PPTX (professional slides for title, sections, content, conclusion), TXT, and ZIP archive bundling

$ Developed interactive chatbot for Q&A based on generated content, with web search fallback, conversation memory, history tracking, and unified LLM client manager for seamless model switching

$ Built quality assurance module with readability analysis using Flesch, Gunning Fog, Coleman-Liau metrics, table summarization from extracted data, and content analyzer for overall rigor

$ Implemented modular design with utility modules for research generation, hierarchical processing, chat management, export handling, configuration, prompts, templates, and error handling (custom exceptions for timeouts, rate limits)