2021.AI

15/05/2025

Meet Ahmed Zewain: Product Data Scientist at 2021. AI 👋

As Product Owner for Grace LLM and Grace MLOps, he drives the roadmap, defines feature requirements, and coordinates data science initiatives. He ensures thorough implementation and testing while fostering collaboration across teams.

Beyond his core responsibilities, Ahmed also engages with clients, presenting products and gathering feedback. He believes this direct interaction helps him better understand client needs and build stronger relationships.

His typical day involves a mix of coordinating with colleagues, writing, reviewing, and coding. He enjoys the casual atmosphere at 2021. AI and the ability to have a direct impact on products.

His path to 2021. AI wasn't always easy. Coming from a refugee family and with limited connections, Ahmed faced challenges finding a job in his field. "It's not easy for someone with my name to hear 'yes, let's hire that guy'," he shares candidly. However, his persistence paid off, leading him to the welcoming and collaborative environment he enjoys today.

Ahmed's advice for those interested in a career in AI? "Be pragmatic. Learn the math and focus on delivering value. Don't just chase the coolest model – find what works and build upon it."

His prediction for the future of LLMs? "Agents and Agentic AI. There will be a rise in ‘black box’ solutions that perform tasks behind the scenes. This will lead to increased demand for tools and regulations to ensure transparency and trust, especially in regions like Europe."

Ahmed is passionate about ethical AI development at 2021. AI. "We're enabling people to think about AI responsibly," he says. "It's important to be on the right side of history."

05/05/2025

Can you keep up with the latest AI models?

As new models emerge and existing ones evolve, your enterprise LLM solutions must be flexible, and be able to adapt to new models - without having to change the user experience.

With our LLM Solution you maintain flexibility over which models you expose to your end users ✅

This means you can curate a LLM experience, tailored to your specific use cases and the users preferences ✨

This approach guarantees that you can always leverage the most advanced capabilities while maintaining a consistent user experience.

Here, Simon Moe Sørensen gives you a brief look into an LLM solution with different models implemented 👇

30/04/2025

Are you aware of all the AI used in your organization? 🤖

If not, it’s risky business.

Shadow AI is emerging as a major challenge in enterprise technology. As companies embrace AI, many unauthorized tools are being implemented and used, creating significant risks. 🔒

While easy access to AI tools encourages employees to take initiative and use AI, it might also expose organizations to vulnerabilities that demand immediate attention.

The main risk factors:
• Security breaches
• Compliance violations
• Potential reputational damage from uncontrolled AI use

We help organizations bring AI out of the shadows and establish effective governance - never stifling innovation but fostering it responsibly. Reach out to learn how 💡

24/04/2025

Stuck in the famous POC loop? 🤯

Many companies get stuck with Proof of Concepts, running experiment after experiment without ever scaling AI solutions for real business impact.

So, how do you move beyond the testing phase and create tangible value? 🤔
In our recent webinar, Dr. Björn Preuß, Ph.D. and Bjarke Arreskov-Hansen shared key strategies to transition from experimentation to ex*****on:

✅ Define a clear purpose
✅ Shift from “lab” to “factory” mindset
✅ Address risks and compliance early

Want to ensure your AI moves beyond POC and creates real business value? Dr. Björn Preuß, Ph.D. gives you the details in his latest blog!

Enjoy the read 👉https://hubs.ly/Q03g4sgR0

15/04/2025

Considering a cloud exit? 🏃☁️

Many companies are shifting from cloud to on-premises due to concerns over data control, security, and compliance. This cloud exit trend reflects a growing reconsideration of LLM deployment strategies, but why?

Key reasons for on-prem:
✅ Full control over sensitive data: On-prem LLMs ensure better data security than cloud solutions
✅ Compliance: On-prem can meet regulations and prevent data from leaving certain jurisdictions
✅ Cost efficiency: Over time, the cost of on-prem LLMs can be comparable to cloud solutions.

User experience remains similar, but the choice depends on your organization’s security and compliance needs 💡

Discover more about AI data security, compliance and on-prem AI solutions 👉 Bjørn Olesen

Peter Sondergaard & Bjarke Arreskov-Hansen 🎙️

09/04/2025

Locally-hosted LLMs: What to consider 🤔

From performance and hardware constraints to governance concerns, choosing the right model isn't just about size — it's about fit.

In our latest blog, we help you cut through the noise with a practical step-by-step guide:

✅ Overview of top open-source models and their strengths
✅ Why self-hosting matters for compliance & control
✅ Understanding the relevant GPU specifications
✅ Trade-offs: parameters, context size, quantization & efficiency

Balance the trade-offs and make the right choice for your AI implementation.

Dive into Guilherme Costa’s and Julian Róin Skovhus’s practical guide now 👉 https://hubs.ly/Q03g4r5K0

19/03/2025

Are you struggling to move beyond the POC stage? 🤔

Establishing a structured approach and a solid technical setup is crucial. For first-timers, thorough preparation is key: plan your production model, define processes, and secure necessary sign-offs.

When executed correctly, you set the stage for scalable growth. 📈

At https://hubs.ly/Q03cgnRT0, we help you navigate this process by providing both expert support and the technical setup you need with our GRACE AI Platform - all to ensure your AI implementation is successful. 🌟

Bjarke Arreskov-Hansen & Dr. Björn Preuß, Ph.D. 🎙️

14/03/2025

There are two ways to adopt LLMs — but only one sets you up for long-term success. 🚀

Here’s why:
Businesses typically follow one of two approaches when introducing large language models.

The first? Deploy a general-purpose AI like ChatGPT or Gemini across the organization and let employees experiment freely.

The second? Take a structured, use-case-driven approach — embedding LLMs into a specific function, like turning static HR documents into a dynamic, chat-based knowledge source.

But here’s the difference: One approach gives more control, compliance, and long-term value. 📈

While the open-door approach might seem like a fast way to introduce AI, it often leads to fragmentation, security concerns, and a lack of alignment with business objectives.

Without oversight, companies risk LLM usage becoming an unstructured experiment rather than a strategic tool.

On the other hand, purpose-driven adoption ensures AI is embedded where it creates real impact — improving efficiency, enhancing decision-making, and driving measurable outcomes. It also allows for governance from day one, ensuring AI systems remain transparent, secure, and scalable. 🔒

At https://hubs.ly/Q03bvbSB0, we help organizations move beyond the hype and into structured, responsible AI implementation — where models don’t just exist in your business but actually work for it.

12/03/2025

AI everywhere. Value nowhere? 🔍

Many organizations rush to expand AI capabilities without first establishing the right strategic foundation, leading to inefficiencies, wasted resources, and missed opportunities.
Think differently, and scale AI in the right way.

In the final webinar of our Common Value Traps in AI series, Björn Preuss and Bjarke Arreskov-Hansen will explore the risks of scaling AI without a clear focus and how to avoid them.

This session will emphasize:
✅ When to build and when to buy - to get that competitive edge
✅ How to avoid the trap of reinventing the wheel when scaling AI
✅ How to maximize efficiency with a scalable AI strategy
Join us to learn how to scale smarter.

📅 March 26, 2025
⏰ 10:00 AM CET
🔗 https://hubs.ly/Q03bt6GN0

18/09/2024

How important is Responsible AI for your business? 💡

In our past EU AI Act webinar, we explored the critical intersection of Responsible AI, the EU AI Act and business success. Here's a glimpse into what the audience, representing a diverse range of industries, had to say:

📑 Importance of Responsible AI:

▶ Nearly half (46%) view Responsible AI as a license to operate, essential for continued business success
▶ 15% see it as a customer requirement
▶ 21% prioritize compliance, while 18% are unsure of its importance

📊 We also asked the participants what they found to be their biggest governance challenge:

▶ Data governance emerged as the biggest challenge for 53% of respondents
▶ Privacy & GDPR came in at 24%
▶ Security & Cyber concerns followed at 15%
▶ Incident/problem management and quality management were significantly lower concerns at 6% and 3%, respectively

The conversation around Responsible AI is gaining momentum, and for good reason. While data governance remains a top challenge, we're committed to helping organizations navigate the complexities of AI implementation 💫

11/09/2024

Discover AI's potential and perils for excellent Customer Experience 🚀

Are you ready to explore how AI can revolutionize your customer service? Join us for an insightful webinar where we'll dive deep into the benefits, challenges and solutions when using AI in customer service.

What you can expect:
✨ Explore AI's game-changing impact on customer service
✨ Learn how to customize AI solutions tailored to your business
✨ Gain insights into the challenges of governing AI and practical solutions to manage them
✨ Walk away with deployable AI tactics to elevate your customer experience

This webinar is designed to equip you with the knowledge and strategies to integrate AI effectively and responsibly into your business practices.

Join us to leverage AI to take your customer experience to the next level 🙌

Are you ready to discover AI's potential and perils for excellent customer experience?Join us for an exploration into how businesses can succeed with AI in...

04/09/2024

How can Large Language Models Revolutionize Knowledge Management? 🔍

AI is transforming how businesses manage and utilize knowledge with Large Language Models (LLMs) at the forefront. These models optimise enterprise knowledge systems, creating a more informed and productive workforce:

An LLM tailored for knowledge management can:

✅ Streamline onboarding
✅ Capture valuable tacit knowledge from experienced employees - and ensure efficient knowledge sharing across the entire company
✅ Tackle information overload, provide precise answers to queries and improve knowledge sharing processes

Read more about how LLMs can revolutionize knowledge management in your company in our latest blog post 👉

Large Language Models offer a revolutionary approach to knowledge management, empowered by AI...