simply-sci

Simply Sci is a utility service that transforms research paper abstracts into clear, accessible language and generates SEO-friendly keywords, enabling authors to share their research more effectively on social platforms and broaden its reach.

Menu

• Tech Stack
• Requirements
• Quickstart
• Project Structure
• API Docs
• Examples
• What's next

Tech Stack

• FastAPI (API framework)
• Uvicorn (ASGI server)
• Poetry (dependency and environment management)

Requirements

• Python >= 3.14
• Dependencies: FastAPI, Uvicorn , Poetry, Pydantic

Quickstart

• Install dependencies
  • poetry install
• Run the API:
  • poetry run dev (this one configured check main.py) OR
  • poetry run uvicorn app.main:app --reload
• Open API docs:
  • http://localhost:8000/docs

Project Structure

• app/api/...: API routers (with versioning)
• app/core/...: config, logging, security.
• app/models/... : db models (future use)
• app/schemas/... : required dto (Data transfer object for ex: req signature object , response object etc)
• app/services/... : common place for keeping business logic
• app/db/...: database base and session (future use).
• migrations/... : db migrations (future use)
• app/middleware/...: common place for middleware
• app/gateways/...: app external connection commom place (open api http call etc)
• docs/...: relevant docs

API Docs

• Swagger UI: http://localhost:8000/docs
• ReDoc: http://localhost:8000/redoc

Examples

how to generate open api token

visit below docs -> click Use this model --> click Create Personal access token

Update token to the below req header to test

github marketplace / gpt 4o model

Quick example from docs/example.md:

curl -X POST 'http://localhost:8000/summary/' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer <YOUR_TOKEN>' \
  -d '{
    "content": "High-precision transmission spectroscopy of temperate super-Earth exoplanets has revealed atmospheric features inconsistent with chemical equilibrium predictions. In this work, we analyze multi-epoch observations obtained with space-based infrared instruments using a hierarchical Bayesian retrieval framework. The inferred molecular abundances suggest the presence of photochemically driven disequilibrium processes, vertical mixing, and possible aerosol formation. While no definitive biosignatures are identified, the observed atmospheric complexity challenges existing models of terrestrial exoplanet evolution."
  }'

Response:

{
	"title": "Studying the Atmospheres of Super-Earth Exoplanets",
	"summary": {
		"problem": "Scientists want to understand the atmospheres of super-Earth exoplanets, which are planets outside our solar system that are larger than Earth.",
		"goal": "The goal is to find out if the chemical makeup of these atmospheres matches what we expect based on scientific models.",
		"what_they_did": "Researchers used advanced space-based tools to observe these planets multiple times and analyzed the data using a special method called Bayesian retrieval.",
		"what_they_found": "They found that the chemical compositions of the atmospheres do not match predictions, suggesting complex processes are happening, like mixing of gases and possible formation of tiny particles.",
		"why_it_matters": "These findings challenge our current understanding of how planets like Earth evolve and could have implications for finding signs of life on other planets."
	},
	"seo_keywords": [
		"super-Earth exoplanets",
		"atmospheric features",
		"chemical equilibrium",
		"Bayesian retrieval",
		"planetary evolution"
	]
}

See more examples in docs/examples.md.

Whats Next

• Add database to store open api req/res
• improve api request time
• understand LLM api attributes
• deploy using ketee (docker+kubernetes)
• Add Vector DB functionality with PG
• In future, connect to LLM model locally and impelmet RAG