ContextChem is a research project exploring how context engineering techniques and large language models (LLMs) can be applied to predict the physicochemical properties of amines.
Instead of treating prediction as a purely data-driven task, ContextChem leverages prompt design, context windows, and retrieval strategies to enable better reasoning in both ranking and regression scenarios.
- Context Engineering: Applying prompt structures, retrieval augmentation, and chaining strategies.
- Chemistry + AI: Bridging contextual modeling with molecular property prediction.
- LangChain Integration: Modular and extensible pipelines for working with LLMs.
Chemical property prediction is traditionally based on costly simulations or experiments. With the right context engineering techniques, LLMs can approximate molecular reasoning and accelerate insights in drug design, material discovery, and industrial chemistry.
git clone https://github.com/yourusername/contextchem.git
cd contextchem
pip install -r requirements.txt