Predict Chemical Reaction
Outcomes Before You React

Chemdem uses AI to predict reaction success and yield before you even start — powered by real squarate-amide data and IBM RXN.

Reaction Input Prediction
+
Product Status : ✓ SUCCESS
Estimated Yield :
78%
Confidence : High · Para-substituted aryl ring detected · Pd/B coupling favored
Why we are building Chemdem

It started with a real chemistry problem, not a trend.

Chemdem began from a challenge inside a university research group. When researchers wanted to make a molecule, they could not easily know in advance whether the reaction would work or what yield to expect. The goal became simple: turn trusted lab knowledge into a tool that helps researchers make better decisions before spending time, reagents, and lab hours.

Research problem
A reaction that might not work — and no way to know the yield in advance.
01
Lab data
Verified reactions from the group's own published papers, structured by hand.
02
Prediction tool
Trusted knowledge turned into an answer you can check before the bench.
03
The founder position

Built between chemistry, code, and design.

Chemdem is being built from a rare position: pharmaceutical science, organic chemistry, software development, and UX design coming together. The aim is not only to make the system technically work, but to make it usable and understandable for chemists at the bench.

Chemistry-first

Designed around real molecules, real reactions, and real research constraints.

Software-built

Structured with modern tools such as Python, Node.js, RDKit, and clean web interfaces.

Made for trust

Every prediction should show its reasoning, confidence, and references.

The first focus

Starting narrow on purpose.

Chemdem's first focus is squaric acid chemistry, specifically amines reacting with diethyl squarate to form squaric acid monoamides. The scope is narrow because trust comes before scale. Before Chemdem expands to more reactions, it must prove the approach on chemistry the research group understands deeply.

Amine
+
Diethyl squarate
EtOH
Squaric acid monoamide
Phase 1 scopeOne reaction family, understood deeply — the foundation everything else is validated against.
The data problem

The real wall is not the algorithm. It is the data.

The hardest part of Chemdem is turning chemistry from papers into reliable machine-readable data. Molecules need to be converted into SMILES, checked, verified, and labelled correctly. A single wrong structure can poison the prediction system, so data quality is treated as the foundation of the whole project.

01
Research paper

Reactions live in published PDFs, drawn as structures.

02
Chemical structure

Each molecule is identified and isolated by hand.

03
SMILES conversion

Structures become machine-readable text strings.

04
Human verification

A chemist confirms every structure is correct.

05
Trusted dataset

Verified, labelled reactions — the foundation.

06
Prediction engine

Only clean data ever reaches the model.

Hover any step to see what happens there.

Transparency

Chemdem is not pretending to be magic.

Chemdem is currently a transparent algorithm, not a fully trained machine-learning model. With a small verified dataset, a neural network would not be the honest tool. Instead, Chemdem reasons by analogy: it compares a new molecule to known reactions and uses the closest real examples to estimate outcome and yield.

“Chemdem does not learn chemistry yet. It remembers it.”
Generic AI hype
Black-box answers
No clear references
Overpromised accuracy
Chemdem approach
Transparent reasoning
Reference-backed predictions
Confidence labels
Known failure detection
How it works

A prediction system that shows its working.

1

Input molecule

The user enters a structure, name, or SMILES.

2

Convert and standardise

The molecule is converted into a clean machine-readable format.

3

Check known failures

If the lab already knows the reaction fails, Chemdem stops and explains why.

4

Find closest reactions

The system compares the molecule with verified reactions in the dataset.

5

Predict by analogy

Yield is estimated from the closest real precedents.

6

Show confidence and references

The output includes confidence, supporting examples, and traceable evidence.

Trust layer

Honesty is the core feature.

A prediction number alone is not enough. Chemdem is designed to show where the answer came from, how confident it is, and when the system is unsure. Chemists can also flag results for human review, helping the dataset improve over time.

Reference-backed predictions
Confidence scoring
Known failure rules
Human review loop
Dataset source labels
Flag-and-improve workflow
Prediction✓ High confidence
Estimated yield74%
SUPPORTING REACTIONS
para-F aniline75%
para-Cl aniline60%
para-Br aniline66%
Roadmap

The future of Chemdem is built in phases.

Phase 1Now

Curated dataset + transparent algorithm

Chemdem is currently focused on verified data, rule-based logic, and explainable prediction.

Phase 2Next

Machine learning when the data is ready

Once the dataset grows into hundreds of verified reactions, Chemdem can move toward a real trained model that learns from trusted chemistry data.

Phase 3Future

Open research contribution network

The long-term goal is a human-curated, referenceable platform where research groups can contribute verified data and improve future predictions.

Open source vision

Open-source, but earned step by step.

Chemdem is open-source in spirit and currently scoped around university research data. The project should not claim universal chemistry prediction too early. It should earn scale through verified contributions, better references, and stronger validation.

Scientist skepticism

Built for people whose job is to doubt the answer.

Chemdem is not here to replace scientific judgement. It is being built for chemists who question every prediction. Their skepticism is not a problem for the project — it is the roadmap. Every challenge makes the platform more traceable, more honest, and more useful.

“You are not building an oracle. You are building something that has to earn trust through evidence.

Chemdem is still being built.

The future depends on better data, stronger validation, and more researchers challenging the system until it earns trust.

Launch App Follow the Research