Blog Post

Pioneering Patient Centric Care Through Advanced Modelling and Linguistics

May 31, 2023

Pioneering Patient Centric Care through Advanced Modelling and Linguistics

At Talking Medicines, we are the world’s gold standard provider of patient intelligence for medications with a vision to improve health outcomes for every patient. As we gather momentum, we continue to grow our capabilities by expanding our Artificial Intelligence (AI) models and methodologies, combining them with our expertise in medicines, statistical principles, and linguistic understanding of how people, patients and those connecting with patients, talk online.

In today’s digital age, an abundance of raw data is available across various online sources, including social media platforms, message boards, and both open and closed online communities. This raw data, although rich in potential insights, is often unstructured and filled with “noise”, meaningless words, making it challenging to derive valuable and actionable information. At Talking Medicines, we understand that different data sources contain unique information, varying in language and structure. Therefore, we meticulously analyse each data source to comprehend its distinctive nature, enabling us to apply our AI-driven tools effectively to structure the data and extract essential information that has high commercial value.

A crucial aspect of our approach is our data classification (focused to patient, healthcare professional), named entity recognition (NER) models provide the “hooks” to extract out data to create populations that can be explore further at a drug, disease, treatment level with an emphasis on linguistic understanding. Our dedicated specialised annotation team works tirelessly to ensure that our AI models are fine-tuned with “gold standard annotation data”, the correctly marked up data that recognizes the nuances in language and context. We recognize that model accuracy alone is not a reliable indicator of real-world performance, especially when dealing with diverse data sources. As a result, our structured data resembles a harmonious network of datasets, each retaining their unique characteristics, rather than a single homogenized dataset that has been reduced to a standardized format. This approach allows us to maintain the richness and diversity of the original data, ultimately leading to more accurate and meaningful data intelligence.

The process of structuring and extracting valuable information from diverse and unstructured data sources is made possible through the utilization of state-of-the-art Artificial Intelligence (AI), Machine Learning (ML), and Natural Language Processing (NLP) methodologies.

Talking Medicine’s Artificial Intelligence Plus Diagram


Artificial Intelligence is a branch of computer science focused on designing systems capable of thinking and acting rationally, emulating human-like intelligence. Machine Learning, a subfield of AI, involves the development of algorithms that learn over time, much like humans, to perform tasks such as prediction, optimization, and decision-making.


ANN & Deep Learning

The advent of Artificial Neural Networks (ANNs) revolutionized AI and ML by attempting to mimic the structure of the human brain, with its complex network of brain cells or neurons. Deep Learning, a more advanced form of ANNs, features a larger number of neurons and a more intricate structure, allowing it to perform complex tasks with higher accuracy, analogous to the human brain.


Transformer & LLMs

Transformers represent a specific version of Deep Learning ANNs, boasting highly desirable technical properties, such as exceptional scalability. Large Language Models (LLMs), which are transformer based ANNs, are trained on vast amounts of natural language data, endowing them with a deep understanding of language. This comprehensive understanding allows LLMs to generate natural language, summarize texts, classify content, and perform various other tasks that involve language processing.


At Talking Medicines, we leverage these cutting-edge technologies to analyse, structure, and derive meaningful insights from the vast and diverse data sources available on social media, ultimately contributing to better patient care and improved health outcomes. Our guiding principles are deeply rooted in a strong foundation of values, including transparency, innovation, community, and quality. We are committed to maintaining the highest standards in ethics, compliance, and governance throughout our operations, ensuring that our AI-driven solutions not only deliver valuable insights but also foster trust and promote responsible practices. “Gold standard” serves as a bedrock for our medicine know-how, AI models and methodologies. This combined with our meticulous attention to data quality enables us to produce reliable and relevant patient intelligence. By upholding these ethical principles and continuously pushing the boundaries of innovation, we are determined to transform the healthcare landscape and make a lasting, positive impact on the lives of patients around the world.


To explore the integration of patient intelligence into your marketing strategy, reach out to us today. Let’s connect and discuss how we can empower your marketing efforts with valuable patient insights

Read More Blogs