In this interview, e27 speaks with Stella Seohyeon Kim, COO and Co-Founder of CoBALT, a company building AI-native systems that help organisations turn everyday interactions into tangible business opportunities. Through its flagship product REALIZER.ai, CoBALT operates at the intersection of sales, business development, and operations, offering a grounded perspective on how AI is being embedded into real workflows as trusted operational infrastructure rather than surface-level features.

This conversation sits within e27’s broader AI coverage, which examines how organisations across the region are building, deploying, and governing AI in practice.

Turning first meetings into real business opportunities

e27: Briefly describe what your organisation does, and where AI plays a meaningful role in your work or offering.

Stella: Cobalt operates REALIZER.ai, an AI-native assistant that turns the people you meet at work into real business opportunities.

Business developers and sales teams meet dozens, sometimes hundreds, of potential customers, partners, and investors through meetings, conferences, and industry events. REALIZER elevates those first encounters from simple contact exchanges into qualified opportunities.

After a meeting, a user can scan a business card, enter an email address, or leave a short voice note about the interaction. From there, Realizer quietly organises the contact, researches the person and company, evaluates the opportunity, and drafts the first follow-up message. The user simply reviews and sends it.

There is a golden window after meeting someone, roughly 48 hours. When meaningful touchpoints are created within that time, the chance of converting the relationship increases dramatically. Realizer is designed to help teams act within that window.

Making individual interactions organisational assets

e27: What is one concrete way AI is currently creating value within your organisation or for your users or customers?

Stella: The greatest value Realizer delivers is turning every individual interaction into a reliable organisational asset.

Instead of relying on personal intuition or fragmented experience, REALIZER enriches and verifies information about prospects, partners, and investors using consistent criteria. It applies a shared logic for evaluating opportunities and recommending next actions.

As a result, teams view opportunities through a common lens, improve pipeline predictability, and move faster without missing critical moments. On an individual level, AI supports not only labour-intensive tasks but also work that requires higher-level reasoning, helping people achieve real outcomes, not just efficiency.

Defining how humans and AI collaborate

e27: What was a key decision or trade-off you had to make when adopting, building, or scaling AI?

Stella: The most difficult, and most important, challenge was defining how humans and AI collaborate.

For effective collaboration, people need to feel confident that they remain in control while still trusting AI-driven decisions. That requires redesigning processes and delivering an experience where AI works almost invisibly, flowing naturally, without users constantly noticing or managing it.

This is the first time in human history that we are working alongside non-human intelligence. There has been trial and error, but our guiding principle is clear. AI should not diminish human value, it should amplify it. Just as electricity became seamlessly embedded into daily life, AI should quietly integrate into workflows and elevate them.

Building trust while managing AI imperfections

e27: Looking back, what has worked better than expected, and what proved more challenging than anticipated?

Stella: Imagine hiring a new employee who executes tasks flawlessly without supervision. That would be ideal. But if you constantly need to double-check their work and clean up mistakes, they quickly become a liability.

AI, especially large language models, is a new kind of junior hire. Depending on how you instruct it, the output can range from excellent to disastrous. It never complains, can repeat tasks endlessly, but it can also hallucinate with complete confidence.

Designing instructions and systems that consistently lead to high-quality outcomes was far more delicate than expected. We believe trust is the foundation of human-AI collaboration, so we built Realizer to earn that trust. It evaluates information across more than 50 sources, applies dozens of validation criteria, and presents not only insights but also confidence levels.

What proved harder was keeping this disciplined AI mostly out of sight, allowing humans to feel effective without constantly confronting AI’s imperfections. AI makes mistakes, just like people do. Managing those failures without burdening users requires a careful balance. It’s challenging, but we believe this balance is what ultimately leads to long-term adoption and genuine affection for the product.

AI requires new ways of working

e27: What is one lesson about applying AI in real-world settings that leaders or founders often underestimate?

Stella: AI is not a magic wand.

Leaders must recognise that adopting AI is not merely a technical upgrade, it is the introduction of a new way of working. No matter how advanced the model is, poorly designed instructions and workflows can make AI worse than useless.

If an organisation fails to adapt how it collaborates with AI, performance may actually decline rather than improve.

Starting small to earn trust

e27: Based on your experience, what is one practical recommendation you would give to organisations that are just starting to explore or scale AI?

Stella: Start small, at a single high-friction decision point.

Rather than pursuing large-scale digital transformation, apply AI to one area where people struggle most or repeatedly waste time. Prove real impact there first, then expand. When there is a clear owner and measurable outcome, AI earns trust and becomes embedded naturally within the organisation.

From AI features to operational infrastructure

e27: Over the next 12 months, how do you expect your organisation’s use of AI, or the role of AI in your industry, to evolve?

Stella: Over the next year, AI will move beyond task-level assistance and become core operational infrastructure.

Within Realizer, AI will increasingly reassess opportunities continuously, monitor signals across channels, and recommend next actions at the team level. Across industries, the competitive edge will shift from having AI features to building trusted, governable AI systems that organisations are willing to rely on in real operations.

Why alignment matters more than speed

e27: Anything else you want to share with the audience?

Stella: The true value of AI is not in making individuals faster, it lies in making organisations more aligned and more decisive.

Working with startups as well as publicly listed Korean companies has made one thing clear. The winners are not the teams with the flashiest models, but those that design AI around trust, clarity, and execution. As AI becomes invisible infrastructure, what matters most is not how impressive it looks, but how deeply and thoughtfully it is integrated.

Stay ahead of how AI is actually being used

This conversation highlights a recurring theme in how AI is moving from experimentation to everyday use. Rather than chasing novelty, CoBALT’s approach centres on trust, alignment, and designing AI that fits naturally into how teams already work. From capturing fleeting first meetings to building shared organisational judgment, Stella Seohyeon Kim’s perspective underscores that the real challenge of AI adoption lies less in models and more in systems, workflows, and human confidence. As AI becomes quieter and more embedded, the organisations that succeed will be those that treat it as operational infrastructure, not a showcase feature.

For more interviews, analysis, and real-world perspectives on how organisations across the region are applying AI in practice, subscribe to our newsletter. You can also explore more AI stories here.

—

Enjoyed this read? Don’t miss out on the next insight. Join our WhatsApp channel for real-time drops.

The e27 team produced this article

We can share your story at e27 too! Engage the Southeast Asian tech ecosystem by bringing your story to the world. You can reach out to us here to get started.

Featured Image Credit: CoBALT