Ciontic.
Explore prototype

IN DEVELOPMENT · PRE-MVP

Cosmetic science,
connected.

From scientific literature to structured evidence.
From individual findings to a knowledge graph.

Ciontic is building a foundation for exploring cosmetic ingredients, their effects, and the science that connects them.

Explore the prototype Follow the evidence

An interactive preview of the direction, using a fixed goldenset example.

ANATOMY OF A CONNECTIONFIG. 01
SCIENTIFIC LITERATURETasić-Kostov et al., 2019Lactobionic acid · in vivo / in vitro
01
has observation
STRUCTURED EVIDENCESkin surface pH decreasedHuman study · 10% w/w formulation · 1 hour
02
INGREDIENTLactobionic acid
CATEGORYSkin barrier
A real goldenset observation, mapped to the proposed graph. A connection preserves context; it does not prove causality.
THE IDEA

A paper is a starting point.
The evidence is the connection.

01 / THE APPROACH

More than finding a paper.
Understanding what it supports.

A search result points to a document. Ciontic is designed to connect the finding to its ingredient, experimental conditions, and source.

  1. 01

    Scientific literature

    Start with a paper. Keep its text, tables, and references connected to the source.

    THE SOURCE
  2. 02

    Structured evidence

    Separate findings into observations, preserving the study model, concentration, and outcome.

    THE FINDING
  3. 03

    Ontology & knowledge graph

    Give ingredients, effects, and biological targets a shared vocabulary and explicit relationships.

    THE CONNECTION
  4. 04

    Evidence-grounded exploration

    Build toward search and reasoning that can trace a response back to the findings behind it.

    THE DIRECTION

Planned workflow · From source literature to traceable exploration.

02 / A REAL EXAMPLE

Small details.
Different conclusions.

A concentration is not a footnote. A human study is not a cell study. And “no significant change” is a finding worth keeping.

This example comes from Ciontic’s manually curated goldenset: reference observations used to shape and evaluate the extraction design.

Inspect the example
GOLDENSET / OBSERVATION 02Direct observation

Lactobionic acid

Lower skin surface pH one hour after application.

Study model
Human
Formulation
10% w/w
Comparison
Placebo & untreated sites
Outcome
Significant decrease

Keep the boundary. In a separate observation, transepidermal water loss (TEWL) showed no significant change after 24-hour occlusion. That is not a demonstrated reduction in water loss.

SOURCE PAPERA 10% Lactobionic acid-containing moisturizer reduces skin surface pH without irritation — An in vivo/in vitro study

Tasić-Kostov, Lukić & Savić · J Cosmet Dermatol · 2019

Paraphrased from the manual goldenset, not a verbatim quotation or a live extraction result.

03 / PRODUCT DIRECTION

Explore the finding.
Keep the source in view.

The prototype carries forward our original interface studies: ingredient profiles, focused evidence cards, and paper-level context.

Browse a fixed example, filter its observations, and follow a finding to its source. It is an interface prototype, with no live search or connected backend.

Open interactive prototype

04 / WHERE THINGS STAND

A foundation in place.
A product in progress.

Ciontic is an independent, bootstrapped product at the pre-MVP stage. The work today is about making the evidence model sound before building on top of it.

IN THE REPOSITORY

Schema, designs & reference examples

PostgreSQL DDL, extraction and graph designs, manual goldensets, and interactive HTML mockups.

STILL TO BUILD

The connected product

The product frontend and backend, automated extraction, graph projection, and evidence-grounded search and reasoning.

LET’S CONNECT

Interested in the science
behind the ingredients?

For research conversations, collaboration, or questions about Ciontic.

founder@ciontic.com