| 🧠 Predictive AI |
| Text Sentiment Analyzer |
Analyzes user-submitted text for sentiment (positive, neutral, negative). |
Logistic Regression or Transformers |
Polarity scoring |
Fast feedback on tone for communication or UX insights. |
| Emotion Classifier |
Detects emotional state in text using neural classifiers. |
Neural networks, optionally Transformers |
Multi-label thresholding |
Enhances emotional understanding in text-based systems. |
| Resume Matcher |
Predicts job fit and tailors resumes accordingly. |
Sentence-BERT, cosine similarity |
ESCO taxonomy, skill extraction |
Optimizes job search success rates. |
| Q&A Mini LLM (RAG) |
Retrieves relevant answers from a mini knowledge base. |
RAG with Sentence-BERT + classifier |
FAISS index, metadata filtering |
Small-scale Q&A testing or embedded knowledge modules. |
| Stock Trader |
Simulates trading strategies and risk-based decisions on market data. |
Quant rules + ML (optional) |
ATR, RSI, backtesting, visualizations |
Explore and evaluate trading logic safely. |
| Self-Trainer |
Enables users to train and use their own classifiers with labeled data. |
User-selected (MLP, SVM, Transformer) |
Data labeling, training interface, model versioning |
Personal AI model builder for classification tasks. |
| 🎨 Generative AI |
| Summarizer |
Creates short summaries of long documents. |
Sumy, T5 |
Chunking, compression tuning |
Speed-read reports, legal docs, articles. |
| Image Captioning |
Describes uploaded images in natural language. |
Vision-to-text (BLIP, similar) |
Image preprocessing, caption ranking |
Improves accessibility and indexing of visuals. |
| Image-to-Story Generator |
Creates a fictional or themed story from a given image. |
FLAN-T5 or similar generative models |
Tone prompts, image captioning |
Fun or educational use for kids, marketing, or entertainment. |
| Rewriter Style |
Changes tone/sentiment of input text. |
LLM-based prompt-to-prompt transfer |
Style profiles, text rewriting templates |
Refines communication for audience or brand tone. |
| Image Editor |
Basic image filters + fashion sketch + cinematic AI editing. |
Hybrid (OpenCV + diffusion-based AI) |
Sketch → render, cinematic look transfer |
Creative exploration for artists and designers. |
| 🤖 AI Agents (Single-Task) |
| AI Detector |
Identifies whether text is AI-generated. |
Transformer classifier |
Token-based heuristics + thresholding |
Detects academic misuse, content fabrication. |
| Fact Checking |
Checks the truth of claims using web evidence. |
RAG-style verifier |
Claim scoring, search & ranking |
Supports responsible information usage. |
| Object Detection |
Finds and labels objects in an uploaded image. |
MobileNet SSD |
Caffe, OpenCV |
Recognize content in photos or assets. |
| 🧠 Agentic AI / Multi-Step Workflows |
| Action Recommender |
Suggests the next best action or task based on user input or summary. |
LLM (instruction-tuned) |
Prompt chaining, task inference |
Assists productivity, PMs, analysts, students. |
| Workflow Orchestrator |
Chains tools and runs multi-step AI pipelines with full history, planner, and logic. |
LLM + conditional chaining |
JSON workflows, session timelines, PAI planner |
Builds and executes complex AI-driven tasks automatically. |
| Prompt Designer |
Guides users through multi-step prompt creation with optional tone, format, and self-critique control. |
LLMs (e.g., FLAN-T5, LLaMA-3) |
Step-based form UI, tone/style inputs, optional self-evaluation pass |
Helps structure effective prompts for any AI task or persona, improving output quality. |
