AI Applications in Pharma and Biotech
Over the last years, the use of Artificial Intelligence in pharma and biotech industry is redefining how scientists develop new drugs, tackle disease and more
Artificial Intelligence is one of the technologies that is expected to impact widely the pharmaceutical sector, not only in its core business but also in its support functions. Pharmaceutical companies are starting to sense that AI can add significant value to the research and development process, improving the early stage of drug development and enabling an informed data analysis on the collected scientific datasets. AI is developing at a rapid rate, and the implementation of this technology can lead to informed decision-making process, increased efficiency of clinical trials, and more optimized innovation process.
When it comes to AI, one of the biggest challenges reported by pharma leaders is the uncertainty on how to structure their data in a way that allows them to enforce machine learning models. Therefore, starting to have an exhaustive vision of where AI can be used in pharma is a useful step towards prioritizing the data to be collected and structured. For that reason, we are listing some possible use cases of AI that can be implemented by pharma companies.
AI and Drug Discovery
In the early process of drug discovery, Machine Learning and Artificial Intelligence have many potential uses, from initial screening of drug compounds to prediction of the success rate of a drug. Additionally, AI can play a big role in identifying drug targets, identifying lead compounds, optimizing their design against multiple property profiles, and identifying synthetic routes to understand the composition of matter. Machine Learning and Artificial Intelligence can make the discovery of new drugs cheaper, quicker and more effective.
Currently, many companies are starting to use AI to identify patterns hidden in large volumes of data. Some of the world's biggest pharmaceutical companies are already investing in disruptive technologies. For example, Pfizer uses machine learning to power its search for immuno-oncology drugs, Sanofi uses an AI platform to hunt for metabolic-disease therapies and Roche is using an AI-powered system to drive the multinational company's search for cancer treatments.
AI and Drug Development
According to a study of Clinical Development Success Rates, 9 out of 10 clinical drugs fail to make it to trials, and a lot fail somewhere between phase I trials and regulatory approval, pushing the costs of drug discovery and development through the roof. On average, the cost to develop new drug can range from $10 million to $2 billion, depending on the type of drug. The R&D process to develop a new drug can be costly, time-consuming, and with high odds to not be put into mass production. What really drives the cost of drug development, is the fact that 90% of medicines that are being tested on people don't reach the market because they end up being unsafe or ineffective.
One of the most common use of machine learning in a company context is predictive analytics, i.e the establishment of accurate predictions and forecasts. In the pharma context, it is increasingly applied to drug development. For example, AI has the potential to transform key steps of clinical trial design from study preparation to execution towards improving trial success rates, thus lowering the pharma R&D burden. Another example is that some companies are successfully implementing AI to predict from a database of molecular structures in advance which potential medicines will work, and which won't.
Apart from improving the efficiency of the drug development process, Artificial Intelligence can bring value in different parts of a company's operations. In the cases below we are going to explain in detail some of these use cases.
AI and Image Analysis
The term "image analysis" is the process of training an algorithm to process pictures, drawings, and videos like humans do. Image analysis can be tailored to specific use cases like logo detection in videos and pictures, scenery analysis in photos, analysis of medical imagery, quality monitoring in factories and more.
Thanks to deep learning, the capabilities of AI in analyzing and understanding what is on an image have multiplied. This particular field of using AI to analyze images is called Computer Vision. We explore some areas where image analysis has potential applications in the pharmaceutical industry: 1.quality control for product and packaging and 2.document analysis for database entry.
Quality Control for Product and Packaging The potential for using AI image processing can range from quantity and condition detection to the inspection of packaging and included items such as printed instructions and dose applicators. Most frequent problems with pharmaceutical quality control and packaging include: counting the number of pills each bottle is filled with, inspecting each pill for accurate size and shape, label validation for product information and barcodes, inspecting packaging for damage and more. Furthermore, companies can use image analyzing to ensure their tracking information is both accurate and instrumented across the channels that would need to access it.
Document Analysis for Database Entry : Another opportunity that arises from image processing is document analysis. For example, documents from past clinical trials can be a valuable source for companies looking to compare all of their clinical trials to decide on best practices. Furthermore, patients reports may contain useful information about a patient's response to a drug and the austerity of any side effects. This can be especially important when a new side effect is discovered or if the patient has an allergy to the drug. Both the pharmaceutical company and its clients can benefit from digitization and centralization of this data, in a sense that the company can aspect every past reaction of the patient and use that information to increase patient's health and gratification.
AI in Human Resources
In an industry where talent identification and talent retention is key, AI offers more and more opportunities for the HR departments. Datasets of potential job candidates, current employees and past candidates paves the way for AI to generate analytical insights into various HR processes. More specifically, AI can play a big role in accumulating value in terms of 1.predicting and preventing employee churn, and in terms of 2.talent acquisition.
Predicting Employee churn: Employee churn, the rate at which employees leave the company, can have a severe impact on firms of all sizes. When an employee leaves your business, it can yield reduction in productivity, additional costs of recruiting and hiring a replacement, and investing time in interviewing and training the new employee. Evidently, not every employee that leaves your company is unsatisfied - after all, some may retire or move to another city. But, if your company is experiencing a high rate of employee churn it may seriously affect your business, and you may want to look into the cause and the prevention of employees churning. Knowing which employee has a high churn rate opens the opportunity for managers to prevent talent attrition by introducing retention strategies. AI-based solutions focusing on churn prediction can help HR teams get insights about the satisfactory level of their employees, and prepare personalized feedback surveys, reward systems and recognition programs that further engage employees.
Predictive Recruitment : AI can deliver value to HR departments in providing analytical insights about candidates. By performing a preliminary analysis of a candidate's resume to cross-check her/his declared skills compared with the skills of successful employees with the same job title, AI can perform as a true matchmaker, speeding up the recruitment process and selecting the best suiting candidates while letting the HR professionals deal with more detailed candidate screening. AI-based solution for predictive recruitment can also eliminate the human bias, by looking only at the set elements, like skills and experience. Hence, leading to more diverse and more productive work environment.
AI in Sales
Sales are very important function of every organization. After all, if companies are not selling, they are not generating any revenue. Increasingly, companies are using AI to improve their bottom line. Competitive sales teams are leveraging Artificial Intelligence as a way to help them scale, increase communication and operate in competitive and global markets. We are going to evaluate some aspects of sales where pharmaceutical companies can leverage AI in order to prevent customer churn and provide personalized recommendations to their customers.
Predicting customer churn : Obtaining new customers can be up to 25 times more expensive than retaining an existing customer. Customer churn can cost businesses inordinate amount of lost revenue every year. With AI based churn prediction, companies can have insights on: what are the drivers of churn, what is their relative importance, and what is their cost. Beyond understanding the reasons for churn on a very granular level, AI enables companies to take action in a data driven way and focus their actions on what will yield results. You can read more about the use case of predicting churn with AI here.
Personalized recommendations : Personalized recommendations has several denominations: predictive marketing, hyper-personalized marketing, individualized marketing etc. It consists in improving customer journey and increasing conversions by showing to customers product recommendations that are relevant to them. Whether a company is operating in B2B or B2C, predictive marketing is becoming more and more common for companies that have large pools of products or customers (either B2B or B2C). By increasing up-selling and cross-selling, AI can help companies increase sales and maximize revenue. You can read more use cases of personalized recommendations here.
AI in Finance
Corporate/Business finance is an important function in any business, since it involves the management of financial resources and activities of the organization. In recent years AI has found innovative applications all across an organization's finance operations. More specifically, Artificial Intelligence can play a big role in generating value in terms of: 1.spend analysis and 2.invoice parsing.
Spend Analysis : Analyzing data takes time and effort. Spend analysis is the process of collecting, cleaning, classifying and analyzing expenditure data with the purpose of decreasing procurement costs and improving efficiency. By using AI for spend analysis, companies can focus their effort on what brings value to the company's finances: identifying and implementing financial optimizations. Pharmaceutical and Biopharmaceutical companies that focus on developing products for several diseases, can leverage AI based spend analysis to help them in optimizing their supplier base and improve process efficiencies. Furthermore, companies can leverage past spend data to predict future costs and adapt prices accordingly. Using spend data, companies can identify price evolution trends and make business decisions accordingly.
Invoice parsing : Being a pharmaceutical company also operating in B2B, vast amounts of invoices are part of everyday operations. Dealing with this amount of invoices is mostly still done manually or semi manually by some companies, which is not optimal in terms of time and efficiency. By training AI to extract data successfully from invoices of any shape and format, CFOs can make room for accountants to focus on other value-creating missions. This process to extract data from invoices is called invoice parsing. Kantify has developed Fyn, an AI solution to automatically detect and classify fields from invoices, regardless of their template, format or language.You can read more about AI invoice parsing here.
AI and Fraud Detection
The rising popularity of online pharmacies opened a new channel that is vulnerable to frauds. Illegal online pharmaceutical sales contribute to billions of dollars in lost revenue to pharmaceutical companies each year. What is even more damaging, is the potential for lost lives from consumers taking counterfeit and fraudulent drugs ordered online. For example, one prescribed drug may be easily purchased online from a pharmacy located in another country for a fraction of the cost, but there is a risk of receiving a counterfeit of the product. AI can help in fighting fraud in online pharmacies, firstly by spotting and identifying websites that sell drugs online. Once the websites are identified, important terms on every page can be identified like names of drugs being sold, the price at which they are sold, dosages etc. This can help in further developing insights from the scraped and analyzed e-website, detecting non authorized vendors and sellers that auction fraudulent drugs online.
Kantify is a team of business and machine learning specialists, building AI models that will help you inflate the growth of your company. We love innovating and placing ideas into practice.
If you are curious in finding out more, feel free to contact us and our team will answer all your questions. You can also book a 15 mn slot to discuss your idea and have a first exchange (in French, English or Dutch).