| AI Detector |
Detects whether text is AI-generated or human-written. |
Transformer-based classifier |
Feature extraction, confidence thresholding |
Helps validate content authenticity or detect AI use in submissions. |
| Emotion Classifier |
Classifies emotional tone of user-submitted text. |
Neural network classifier (text2emotion, optionally Transformers) |
Threshold-based multi-emotion tagging |
Enhances sentiment awareness in text communications. |
| Fact Checking |
Evaluates the truthfulness of factual claims using search-based evidence. |
Retrieval-Augmented Verification |
Web scraping, keyword extraction, scoring logic |
Supports misinformation detection and claim verification. |
| Image Captioning |
Generates descriptive captions for uploaded images. |
Vision-to-Text Model (e.g., BLIP) |
Image preprocessing, caption ranking |
Improves accessibility and image documentation. |
| Image Editor |
Allows basic editing of uploaded images and supports advanced AI-enhanced conversions. |
Hybrid editing (traditional filters + AI diffusion/style transfer) |
OpenCV/Pillow for standard filters, plus text/image prompt‑based sketch↔image, style‑transfer modules |
Empowers designers and creators: from quick edits to fashion‑design mockups and cinematic effects, all in one tool. |
| Image-to-Story Generator |
Turns an uploaded image into a fictional or descriptive story. |
FLAN-T5, Transformer-based generation |
Tone control, prompt templating |
Creative storytelling, engagement, or educational use. |
| Object Detection |
Detects and labels objects in uploaded images. |
MobileNetSSD (Caffe model) |
OpenCV DNN |
Useful for security, tagging, and content recognition tasks. |
| Game Theory (Poker Solver) |
GTO strategy solver using Counterfactual Regret Minimization (CFR). |
Reinforcement Learning, Game Theory (CFR) |
Hand abstraction, Game tree traversal, Strategy caching |
Valuable for effective decision-making, training, or AI benchmarking. |
| Q&A Mini LLM (RAG) |
Answers questions using a hybrid RAG approach over a toy dataset. |
Sentence-BERT + Logistic Regression |
FAISS vector index, metadata filtering |
Demonstrates scalable Q&A logic; will evolve into MBTI personality system. |
| Summarizer |
Summarizes long documents into concise versions. |
Extractive (Sumy) or Abstractive (T5) summarizers |
Text chunking, compression tuning |
Saves time on document reviews, legal or academic use. |
| Text Sentiment Analyzer |
Analyzes positive/negative/neutral sentiment in user text. |
Logistic regression or shallow NN (or Transformers optionally) |
Polarity scoring |
Quick feedback on user or customer tone. |
| Resume Matcher |
Matches a user’s resume to a job description, scores the match, and tailors the resume. |
Sentence-BERT, Cosine Similarity |
ESCO taxonomy-based skill extraction, semantic matching, resume restructuring |
Helps improve job application success by aligning resumes with employer needs. |
| Self-Trainer |
Allows users to upload, label, and train their own models, then run inference interactively. |
User-selected ML algorithms (e.g., MLP, Logistic Regression, or Transformer) |
Custom dataset labeling, local or cloud-based model storage, training interface, evaluation metrics |
Empowers users to build personalized AI models with full lifecycle: data → training → prediction. |
| Action Recommender |
Suggests next steps, risks, or priorities based on input project summaries or discussions. |
Instruction-tuned LLM (e.g., FLAN-T5, GPT-style models) |
Prompt templating, context mapping |
Helps PMs and analysts turn text into actionable steps. |
| Workflow Orchestrator |
Executes multi-step AI workflows with visual timeline, chaining multiple tools (e.g., detect → summarize → recommend). |
LLM-based decision engine + conditional logic + subtool chaining |
FastAPI + JSON timelines + session history + tool planner |
Allows chaining tools into intelligent automation flows for productivity, analysis, or narrative generation. |
| Rewriter Style |
Rewrites text into different tones (professional, warm, excited, etc.) and sentiments (neutral, positive, negative). |
LLM-based style transfer (e.g., FLAN-T5 or GPT variants) |
Prompt design, tone profiles, and form-based rewriting interface |
Ideal for tailoring communication (emails, marketing, resumes) to match the desired audience or mood. |
