Lessons from a Long-Term Strategic Partnership

Based on years of collaboration, this article highlights the strategic partnership between Organización Corona, a Colombian multinational specializing in home improvement, and the Data Analytics and Strategic Business Intelligence program at Long Island University (LIU). 

This mutually beneficial relationship allows LIU students to tackle real-world projects from Corona, ranging from machine learning forecasting models to artificial intelligence (AI) chatbots. These initiatives provide students with high-level technical experience while delivering actionable business solutions. 

The following interview captures the perspectives of Rodrigo Estrada (director of innovation, Corona), Juan F. Monsalvo (lead data scientist, Corona) and Associate Professor Juan R. Jaramillo (LIU) on the success of this academic-industry synergy.

Interview with the Director of Innovation 

In a fast-paced industry, why is it worth the company’s time to invest a full semester into a student collaboration? 

Rodrigo Estrada: For us, this investment of time goes beyond simple collaboration; it is about building a three-way shared value ecosystem. We view this as a strategic symbiosis between the university, the student and the company.  First, we provide students with real, current industry problems, allowing them to apply their knowledge in live scenarios – something irreplaceable for their development as the talent of the future. Second, for the university, it is a critical opportunity to validate the relevance of its curricula. And third, for Corona, it is a strategic commitment: it allows us to see what’s happening “out there” and build internal capabilities regarding new technologies, ensuring our future relevance.

How does this partnership function as an “external R&D lab” for exploring ideas that your internal roadmap can’t prioritize?

Estrada: This alliance functions as a “safe harbor” for uncertainty. In day-to-day corporate operations, resources are tied to short-term goals and an immediate, tangible return on investment (ROI). However, disruptive innovation requires oxygen and freedom to fail.  By collaborating with academia, we create an environment in which the primary objective shifts from profitability to learning and capability building. It gives us access to fresh minds with the time and intellectual curiosity to explore emerging technologies and unexplored paths that our internal roadmap could not financially justify.  

Furthermore, we gain what we call cognitive diversity: We break the company’s internal bias by incorporating the perspective of young students, often from different cultures and countries, who tackle problems without the usual corporate preconceptions. 

As we often say at Corona, “What the company can do depends on what the company knows”; this initiative expands that frontier of knowledge.

When you sit in those final presentations, what specific qualities are you looking for in a student’s delivery and solution?

Estrada: Beyond the technical solution, we seek a mindset. In these final presentations, we evaluate four fundamental pillars that distinguish a good executor from a future leader:

1. Problem Redefinition: The critical capacity to understand, question and, if necessary, redefine the problem to address the root cause. 
2. Capacity for Synthesis: The ability to connect scattered dots between the problem and the solution and communicate them in a concise and practical manner. 
3. Human Skills (Soft Skills): Empathy, leadership and assertive communication are just as vital as technical knowledge. 
4. Critical Thinking and Fundamentals: Solid foundation and the ability to demonstrate “learning how to learn” and constant reinvention.

How does this collaboration help the company stay at the forefront of innovation compared to competitors who don’t collaborate with academia?

Estrada: The competitive advantage lies in the culture of continuous learning and reskilling, which these programs inject into our organization. By interacting with academia, our internal teams are challenged to stay current and look beyond their day-to-day operations. It grants us capacity for foresight: helping us see and understand technological futures before they become industry standards, allowing us to prepare our internal capabilities while the technology matures and costs are optimized. Fundamentally, it is a tool to ensure Corona’s relevance and survival in the long term.

Beyond the technical results, how does Corona view its role in giving back to the academic community and helping shape the next generation of data professionals?

Estrada: We view this as an inescapable ethical and social responsibility. As an organization grows and receives value from society, it has a duty to give back. Our philanthropic role here is to act as a facilitating bridge. We aim to smooth the learning curve and ease the students’ transition into the workforce by exposing them to corporate realities before they graduate. At the same time, we help universities remain relevant by providing feedback on the real challenges we face, bridging the gap between academic theory and industrial practice. However, there is also valuable introspection in this act: By helping shape the next generation, we help Corona recognize its imperative need to transform. It is a way of generating social value that, ultimately, strengthens the industrial fabric upon which we all depend. 

Interview with the Data Scientist 

How do you select which internal projects are “student-ready” versus those that stay strictly in-house?

Juan F. Monsalvo: The selection is based on a key distinction: exploration versus exploitation.

For students, we select projects of high uncertainty, where the main objective is not immediate efficiency but validating technical feasibility and developing new capabilities. These are challenges designed to navigate different technological paths and understand which one holds the greatest potential. We look for the student to act as an explorer, helping us to “see and understand futures,” preparing us for the moment when those technologies mature and their costs are reduced.  

Conversely, the projects we keep strictly in-house are those in which the technology, value capture and solutions are clear and predictable and therefore must be executed with speed and operational efficiency.

What has been the most surprising technical solution or approach a student team has proposed that your internal team hadn’t considered?

Monsalvo: We had a revealing case in a sentiment analysis project for product reviews. The student team proposed a preprocessing layer using a large language model (LLM) before applying sentiment analysis. 

The LLM was tasked with breaking down and segmenting the text to identify which parts specifically referred to the product, point-of-sale service and delivery logistics. This was brilliant because we discovered that more than 40% of our reviews contained mixed information; a customer could love the product (positive) but hate the delivery (negative). Previously, this might have been erroneously classified as a general negative product review. Thanks to this granular approach, we managed to better understand the opportunities for improvement.

Data science is 80% data cleaning. How do you prepare students for the “messiness” of your company’s actual data?

Monsalvo: Our philosophy is: zero isolation from reality. We provide students with raw, real company data, complete with all its imperfections, noise and gaps. We believe that providing “clean” or synthetic data would do a disservice to the students professional preparation.

However, we don’t leave them alone in the chaos. Because we have already scrubbed this data internally, we know the traps and necessary actions. We act as guides: Instead of telling them what to clean, we pose challenges and strategic questions that highlight inconsistencies. We aim to trigger their critical thinking, leading them to question: Is this data useful? Does this correlation make sense? We want the students to learn that data cleaning is not an operational task but a fundamental analytical process. 

Beyond the final code, how does interacting with these students benefit your team’s technical culture or your approach to problem-solving?

Monsalvo: The most valuable technical ROI is mental openness. Interacting with these teams exposes us to a cognitive diversity that breaks our corporate tunnel vision. Students – lacking company biases – 
explore alternative paths and ask fundamental questions that we sometimes take for granted.

Additionally, there is a benefit of technical reciprocity: To guide students who have fresh knowledge, our internal team is forced to update and renew its own knowledge. It challenges us to be better technical interlocutors and keep up with the innovative solutions they propose. It is an exercise that raises the technical bar for the entire organization.

Interview with the Professor

What sparked the initial conversation with Corona, and why did you feel their projects were a good fit for your curriculum?

Juan R. Jaramillo: I am a strong advocate for experiential learning; real-world projects provide students with a motivation level that a textbook simply cannot replicate. I’ve observed that when external partners enter the classroom, student engagement and attitude improve significantly. Previously, I worked with entities that provided single-semester problems. Although valuable, these short-term projects had two main drawbacks: Organizations expected immediate solutions that weren’t always feasible, and the constant rotation of partners made it difficult to build meaningful momentum. 

The spark for this specific partnership happened in 2018. While chairing the INFORMS Analytics Society’s Innovative Applications in Analytics Award (IAAA), I invited Rodrigo Estrada to join our judging panel. During our discussions, he mentioned Corona’s interest in creating a dedicated space to evolve their AI and analytics capabilities. I shared my vision for a long-term academic partnership with LIU. We quickly realized our goals were aligned to foster sustainable growth for both the company and our students.

How does the relationship work?

Jaramillo: From the outset, we eliminated financial ties and focused on prioritizing the learning process over immediate commercial benefit. This framework provides the flexibility necessary for academic exploration and relieves student pressure when results don’t meet corporate expectations.

During this time, members of both organizations have developed a deep professional rapport. Our dialogue has evolved beyond simply sourcing semester projects; we now engage in a genuine exchange of ideas. We introduce Corona to emerging AI and machine learning methodologies, and they provide us with invaluable insights into how these theories are implemented in a complex corporate environment. This synergy has made project selection seamless; Corona understands our students’ profiles and identifies challenges that perfectly align with our curriculum.

How does working on a live corporate project change the students’ engagement compared to using clean or synthetic datasets?

Jaramillo: Working with live corporate projects introduces a level of complexity that synthetic datasets simply cannot match. Real-world data is inherently unstructured, messy and incomplete; navigating these “unclean” datasets is a vital, challenging skill that most students are experiencing for the first time. 

Beyond the technical challenge, the level of engagement significantly increases when the work has the potential for real-world impact. Our students meet with Corona’s data scientists biweekly – approximately eight times a semester. This creates a shift in accountability: students often feel a greater sense of responsibility to deliver quality work to a professional peer.   Furthermore, we foster a culture of “collaborative competition.” By allowing students to view each other’s progress, they are inspired to learn from their peers while striving to develop unique, high-value solutions.

How has the collaboration evolved over the semesters to keep up with the fast-moving pace of data science?

Jaramillo: Our collaboration coincided with the rise of transformer models and the integration of generative AI into the industry. This timing allowed us to create a rapid innovation cycle: On the academic side, we identify emerging techniques and algorithms, which Corona then reviews and tests within their production environments.

The relationship has since evolved beyond the scope of individual projects. We now meet periodically to explore broader ideas and strategic opportunities. These sessions help Corona implement innovative solutions while allowing LIU to continuously update our curriculum and ensure our teaching remains relevant.

Conclusions

The collaboration between Organización Corona and LIU demonstrates that a long-term academic-corporate partnership creates a unique innovation ecosystem. By prioritizing exploration over immediate profit, Corona gains fresh technical insights and cognitive diversity, and students develop essential professional capabilities by solving messy, real-world data challenges in a fast-evolving industrial environment.