ALCHE: Optimizing Pharmaceutical Manufacturing with AI, Moving Beyond Human ‘Intuition’

- ALCHE has launched "PRIME," an AI-driven SaaS platform that optimizes the complex pharmaceutical manufacturing process by proposing ideal chemical reactions, temperatures, and mixing ratios.
- Unlike traditional methods that rely on expert "gut feeling," PRIME utilizes Explainable AI to provide clear scientific evidence for its recommendations, ensuring researchers can trust and verify the data-driven process while complying with strict safety regulations.
- Founded by CEO Dong-ju Kang, an engineering Ph.D. from Seoul National University, the startup aims to expand from chemical synthetic drugs into the biopharmaceutical sector, setting a new global standard for efficient and predictable drug production.

Dong-ju Kang, CEO of ALCHE/ source=IT DongA

Developing a single new drug requires an astronomical amount of time and cost. Consequently, the adoption of Artificial Intelligence (AI) in the pharmaceutical industry has primarily focused on identifying drug candidates. However, even if a candidate substance is discovered with great difficulty, how to produce it stably and efficiently at scale is a problem on an entirely different dimension.

Traditional designs for pharmaceutical manufacturing processes have often relied on the experience, know-how, and so-called ‘gut feeling’ of experts. This process frequently led to immense costs due to repeated trial and error. For instance, unpredictable impurities would sometimes arise, threatening patient safety or causing massive losses for pharmaceutical companies. Ultimately, the method of manufacturing drugs safely and efficiently while complying with regulations is just as much a core competitive advantage for the pharmaceutical industry as finding a good medicine.

ALCHE (CEO Dong-ju Kang) is a deeptech startup aiming to solve these chronic manufacturing and quality control issues in the pharmaceutical industry through AI. Their developed pharmaceutical lifecycle quality management SaaS(Software as a Service) solution, PRIME, is not a simple statistical tool. The AI itself designs and proposes the optimal process, ranging from the selection of raw materials to chemical reactions and additive mixing ratios.

Founded in March 2024, ALCHE, led by CEO Kang—a process AI expert with a Ph.D. in engineering from Seoul National University—is transforming the experience-dependent pharmaceutical manufacturing process into a data-driven, predictable science. We met with Kang, who has ventured into the new frontier of manufacturing process optimization, to hear about the vision of ALCHE.

- You have a Ph.D. in Chemical and Biomolecular Engineering. What motivated you to start a business by applying AI to the manufacturing process rather than drug development?
: I majored in Chemical and Biomolecular Engineering at Seoul National University and went on to complete my doctoral degree. I was originally researching traditional chemical engineering, but around 2016, when the AlphaGo boom occurred, I felt a great interest in AI technology. I thought that applying this innovative AI technology to the fields of chemical engineering and pharmaceuticals would create tremendous synergy. Furthermore, while researching substances with new properties and new medicines during my doctoral studies, I realized that the manufacturing process is just as important as the drug development itself.

Generally, even when the pharmaceutical industry utilizes AI, they only focus on what kind of drug to develop with it. If drug development is like drawing an excellent blueprint, I wanted to devise a way to build the structure most efficiently and sturdily based on that blueprint. Rather than simply following the trends everyone else is pursuing, I founded ALCHE to pioneer a new trend in pharmaceutical process optimization, a problem no one has properly solved.

PRIME, ALCHE's solution for optimizing pharmaceutical manufacturing processes / source=ALCHE

- What kind of solution is PRIME, ALCHE’s core product, specifically?
: In short, PRIME is an AI platform that optimizes and rationalizes the entire pharmaceutical manufacturing process. The process of making medicine is chemically very complex. For example, an API(Active Pharmaceutical Ingredient) like Acetylsalicylic Acid or Sildenafi lis extracted through a chemical reaction from specific raw materials. Then, various additives must be mixed, and it must go through a complex process before a finished drug like Aspirin or Viagra, which we are familiar with, is born.

In this series of processes, there are countless variables, from which raw materials to select to the pressure and temperature during processing, and the mixing ratios of numerous additives. Additionally, very strict regulations follow to control various impurities that inevitably occur during this process. PRIME is a system where AI, having learned from vast amounts of data, first proposes the most ideal pressure, temperature, and additive ratios to researchers while complying with all these demanding regulations. Researchers can maximize the perfection of the medicine by conducting experiments based on the optimal recipe suggested by the AI.

- There must have been existing software related to processes in the pharmaceutical market. What is the decisive differentiation of PRIME?
: In the past, people had to go through numerous trials and errors because they relied on the personal intuition and "feeling" of experts or simple statistical data. I proudly claim that we are virtually the first to offer a solution like PRIME, where AI actively analyzes chemical processes and preemptively proposes optimal manufacturing conditions.

Above all, the most powerful weapon of PRIME is the application of Explainable AI technology. While conventional AI simply threw out a result saying "this condition is good," PRIME shows the clear causal relationship—why these results were derived and why this temperature and mixing ratio are the optimal process conditions. Researchers in the pharmaceutical industry, who deal with human lives, cannot easily accept results from even the most outstanding AI if they don't know the basis. Because PRIME clearly presents that scientific evidence, it helps researchers fully trust the results and make next-step decisions with confidence.

Through PRIME, pharmaceutical manufacturing processes, including temperature and mixing ratios, can be optimized / source=ALCHE

- What business value can a company gain when they adopt PRIME?
: PRIME is provided in the form of a cloud-based B2B SaaS. Fundamentally, the pharmaceutical business is an IP(Intellectual Property) business, and the unique process know-how and data held by each company are its lifeblood.

Newly established pharmaceutical companies or bio-ventures that are just starting out lack the field experience and data to produce a substance even if they find a great one. They can compensate for that gap through the PRIMEAI engine to establish a production system quickly and stably. Conversely, traditional pharmaceutical companies with long histories can load and train their vast accumulated data onto PRIME. This makes it possible to advance it into a more powerful, proprietary production solution perfectly customized for the company. Regardless of the company's size, we become an essential partner that dramatically reduces development periods and failure costs.

- What is the actual market response and the current status of the business?
: We launched the beta version of PRIME in January of this year. We are currently actively conducting PoC(Proof of Concept) to apply the solution to actual sites together with mid-sized pharmaceutical companies. Since it is a solution that did not exist in the market before, the pharmaceutical industry is giving positive feedback, calling it a very fresh and practical approach.

Recognizing this potential, we were recently selected for a national project by the Ministry of Trade, Industry and Energy(MOTIE) and secured funding for research, development, and commercialization. Based on this, we are further refining the solution and accelerating full-scale commercialization

- As a technology-based startup, how did the Initial Startup Package program from the Seoul National University of Science and Technology (SeoulTech) Startup Support Foundation help?
: Having spent a long time as a researcher, I had solid technology and ideas, but I was lacking in many aspects regarding how to actually run a company as a business. The program from SeoulTech went beyond simple financial support, providing various one-on-one mentoring and educational seminars necessary for overall corporate operation, such as BM(Business Model), IR(Investor Relations), and global business expansion.

During a time when we were only running forward frantically, I was able to look back at the overall business from an essential perspective to see if our direction was correct. Because it played a decisive role in rebuilding the company's framework sturdily from the basics, I highly recommend this program to other startups that have great technology but find business operations overwhelming.

Dong-ju Kang, CEO of ALCHE/ source=IT DongA

- What is the ultimate vision and goal of ALCHE?
: It is to innovate the paradigm in the fields of pharmaceutical quality control and process design. Currently, we are focusing on the field of chemical synthetic drugs, but in the future, we intend to expand our solution to the biopharmaceutical field and establish ourselves as a global solution that becomes the standard for all pharmaceutical development.

Through ALCHE's PRIME, pharmaceutical companies will be able to alleviate the immense losses and anxieties caused by impurity issues, and patients will be provided with safer and more reliable medicines. I hope you will watch with interest the journey of ALCHE as it creates a new standard for pharmaceutical quality control in the global pharmaceutical market.

By Young-woo Kim (pengo@itdonga.com)