Synergy over Silos – the New Age of Knowledge

I recently sat down with Gabriel “Gabe” Karawani, co-founder of ClearPeople, for an episode of our Polaris podcast. If you’ve ever felt the frustration of searching across countless apps for that one piece of information, this conversation is for you. In today’s fast-paced digital workplace, information is everywhere – and that’s exactly the problem. We have data in emails, chat threads, SharePoint sites, CRMs, you name it. Gabe and I explored how integrating knowledge with AI can turn this information overload from a burden into a competitive advantage, and the insights were both eye-opening and practical. 

The High Cost of Scattered Knowledge 

Right at the start, we tackled a striking statistic: the average employee switches between applications and websites about 1,200 times in a day, losing almost four hours a week just reorienting themselves after each switch. It’s what experts now call the “toggle tax.” Think about that – almost 10% of our work time, gone. If you’re a business leader, you feel this in lost productivity and employee burnout. It’s not that people aren’t working hard; it’s that our tools aren’t working together.

The first generation of workplace AI assistants didn’t fully solve this either. In many cases, they added yet another silo. Gabe called them “islands of AI.” You might have an AI helping with your sales CRM, another embedded in your intranet, and perhaps your team dabbled with ChatGPT. None of these talked to each other. The result? If you had to prep for a client meeting, you’d query your Microsoft 365 Copilot for internal files and separately check Salesforce for client data, and dig into ServiceNow for support tickets. It’s exhausting – and inefficient. 

Omni AI: Connecting the Dots 

What business leaders are asking for now is pretty simple: Can our AI systems please work together? This emerging approach even has a name, one we’ve been using at Synozur: Omni AI. The idea is an AI that can draw from all your relevant data, not just one source, and provide a unified answer. Instead of multiple isolated “smart” tools, you get one integrated intelligence that spans your organization. I wrote about this trend earlier, and since then the industry is moving fast. In fact, new open standards are coming to life – we mentioned Agent-to-Agent (A2A) communication and the Model Context Protocol (MCP). That’s where we’re headed, and 2025 is shaping up to be a landmark year for it. 

Meet Atlas: A Knowledge Compass 

ClearPeople’s platform, Atlas, addresses these issues by intelligently connecting and classifying knowledge. For example, Atlas tags documents with location and department, ensuring employees get the right answer based on context. This prevents mix-ups, such as confusing UK and US policies. Experts can easily contribute knowledge, making it accessible enterprise-wide and fostering a learning culture.

Gabe emphasized that AI must use proprietary, high-quality company data to deliver real value; generic AI produces generic results. According to Synozur research, US businesses could lose $280 billion in productivity this year due to inefficient knowledge searches—an opportunity for integrated AI to reclaim lost time.

Security and transparency are critical. Atlas cites sources, allows user feedback, and aligns with standards like ISO 42001 and the EU AI Act, ensuring trustworthy and compliant AI use. Data governance and responsible deployment are essential to avoid missteps or accidental data exposure.

Success is measured by metrics like faster searches and better onboarding. One law firm reduced search times tenfold, while another improved customer response times. These gains translate into higher productivity and client satisfaction.

Conclusion: Partnership Between People, Knowledge, and AI 

By the end of the episode, one thing was crystal clear: the future belongs to companies that break down knowledge silos. Whether through platforms like Atlas or similar integrated approaches, the goal is to let employees spend their brainpower on creativity and decision-making – not on fruitless searches or duplicating work that’s already been done elsewhere in the org. It’s exciting -we’re moving from an era of isolated information and one-size-fits-all AI to an era of context-rich, collaborative intelligence. 

As you navigate your own organization’s AI and knowledge strategy, think about the gaps that might not be obvious at first. Where are teams reinventing the wheel? Which “expert in X” keeps getting pinged for answers that could be documented? And how many different places does your company store “truth” for a given topic? Tackling these questions with the right tools – and yes, a bit of optimism – can yield tremendous outcomes. 

Check out our latest episode of Polaris to hear even more from Gabe Karawani.

Polaris is available on Apple, Spotify, Amazon, YouTube or wherever you get your favorite podcasts. Thanks.

Show Notes

Key Takeaways 

The “Toggle Tax” is Real: Fragmented apps and data silos cause employees to toggle between tools ~1,200 times a day, wasting nearly 4 hours a week (about 9% of work time) just reorienting ¹. Unified knowledge systems can recover this lost productivity. 

First-Gen AI Fell Short: Early AI assistants created isolated “islands of AI” within each tool, adding to silos. Business leaders now demand Omni AI – solutions that integrate data across all platforms for cohesive answers. 

Integrated Knowledge = Competitive Edge: Companies relying only on public AI (like generic chatbots) get “vanilla” answers, while those who leverage proprietary knowledge via tools like Atlas gain trusted, context-aware insights tailored to their business – a significant competitive advantage. 

Atlas Platform in Action: Atlas, ClearPeople’s intelligent knowledge platform, auto-captures and classifies information across Microsoft 365 and other systems, acting as an “ultimate knowledge assistant” that delivers precise answers with sources cited for trust. This reduces search time and surfaces expertise within the organization. 

AI Needs Quality Data: Generative AI isn’t magic – it’s only as smart as the data you feed it. Context is king – AI must know which knowledge base or document version is relevant. A key Atlas benefit is robust metadata (like tagging content by client, project, or jurisdiction) so AI uses the right information when answering. 

Ethical, Responsible AI is Essential: Enterprises should implement AI with transparency, security, and governance. Gabe notes the importance of ethical AI use, complying with standards (like ISO/IEC 42001 AI management systems) and emerging regulations (EU AI Act) while building a culture that encourages sharing knowledge responsibly. 

People + AI, Not vs. AI: Successful adoption requires empowering people with AI, not replacing them. Atlas’s approach augments human expertise – e.g. capturing experts’ know-how and making it discoverable – so employees spend less time searching and more time innovating and collaborating on higher-value tasks. 

Sound Bites (Gabe Karawani) 

• “It became a case of ‘The answer is GenAI – now what’s the question?’ People assumed generative AI would magically solve knowledge problems. In reality, AI needs guidance and context to be useful.” 

• “If you run your business on public AI, you’ll get vanilla answers. Your competitor using AI trained on their own knowledge will eat your lunch with richer, precise insights.” 

• “Think of Atlas as your ultimate knowledge assistant. It connects you instantly with the right information and experts inside your organization by automatically capturing and classifying knowledge at scale.” 

• “Unless an AI agent can connect to the right contextual knowledge, it’s likely to give a perfectly credible answer based on the wrong data. That’s a big problem we solve – making sure AI is grounded in truth.” 

References

Toggle Tax: Harvard Business Review – “How Much Time and Energy Do We Waste Toggling Between Applications?” (Murty et al., 2022). Study of 137 workers found ~1,200 app/website toggles per day and ~4 hours per week lost to reorienting. https://hbr.org/2022/08/how-much-time-and-energy-do-we-waste-toggling-between-applications  

Retirement of Viva Topics (February 2024) to refocus on generative AI, with full retirement by Feb 2025. Organizations using Viva Topics are advised to prepare alternatives for knowledge discovery. https://www.reworked.co/digital-workplace/microsoft-is-retiring-viva-topics-heres-what-you-can-do/  

Agent2Agent (A2A) Protocol – An open standard (initially by Microsoft and Google, now Linux Foundation) for secure, interoperable communication between AI agents. Announced in 2025 and backed by over 100 tech companies, A2A enables different AI systems to “talk” to each other and collaborate. https://www.zdnet.com/article/linux-foundation-adopts-a2a-protocol-to-help-solve-one-of-ais-most-pressing-challenges/  

Model Context Protocol (MCP) – An emerging open standard (led by Anthropic) that provides a universal way for an AI agent to connect to external tools, APIs, and data sources ^2 2. MCP acts like a “USB-C port for AI,” complementing A2A by grounding agents in shared data. https://www.zdnet.com/article/linux-foundation-adopts-a2a-protocol-to-help-solve-one-of-ais-most-pressing-challenges/  

ISO/IEC 42001:2023 – The first international standard for AI Management Systems (AIMS), published Dec 2023. It provides a framework for governing AI use (risk management, transparency, bias mitigation, etc.) to ensure responsible, ethical AI deployment https://www.iso.org/standard/42001

EU AI Act – The European Union’s Artificial Intelligence Act (entered into force Aug 2024) is the world’s first comprehensive AI law. It introduces a risk-based framework restricting certain AI practices, imposing transparency for AI systems, and requiring extra compliance for “high-risk” AI (e.g. in hiring or biometric ID). Artificial intelligence act

Guest Information

Gabe Karawani - LinkedIn

ClearPeople – A digital solutions company co-founded by Gabe Karawani, specializing in knowledge management on Microsoft 365. Creator of the Atlas Intelligent Knowledge Platform. https://www.clearpeople.com

Atlas Intelligent Knowledge Platform – ClearPeople’s AI-powered knowledge platform for enterprises. It centralizes and auto-classifies content across Microsoft 365 (Teams, SharePoint, etc.) and other sources, providing a single trusted hub for knowledge discovery, enterprise search, and AI assistance. https://www.clearpeople.com/atlas-platform

Events

M365 Community Days NYC | July 25, 2025 at Microsoft NYC in Times Square

Seattle Tech Week |  July 28 - August 1, Seattle WA

TechCon 365 Atlanta | August 11-15 at the Georgia World Congress Center.

FY26 Forward - Microsoft Partner Success Strategies - August 26. I’ll be sitting down with our CEO Michelle Caldwell for a fireside chat about Microsoft's plans for its new fiscal year and how Microsoft partners can align for success.

North American Collaboration Summit | September 7-9 Branson. MO

TribalNet Conference | September 14-18, 2025, Reno Nevada

Vancouver AI Summit | October 20, 2025 Vancouver BC

TechCon 365 Dallas - | November 3-7, 2025 (Dallas, TX). Irving Convention Center

Microsoft Ignite 2025 – Nov 17-21, 2025 (San Francisco, CA). 

ESPC25 (European SharePoint Conference 2025) | Dec 1-4, 2025 (Convention Centre Dublin (CCD) Dublin, Ireland). 

Production

Polaris is a production of Synozur – the transformation company. Synozur reimagines business for our clients, navigating the complexities of transformation and strategy with ease.

Polaris is produced with the help of our friends at Riverside.fm. Our theme song, “Alternative Dream” is provided courtesy of Adobe.  Additional music and sound provided by IndieGuy Records. Graphic design by Josh Brantley.

Chapters

00:00 Introduction to Gabriel Karawani and Clear People

06:21 The Evolution of Atlas: From Services to Product

09:50 Navigating the AI Landscape: Misconceptions and Realities

15:34 Success Stories: Real-World Applications of Atlas

21:23 The Importance of Tailored AI Solutions

26:18 Ethical Considerations in AI Implementation

29:16 Measuring Success: Key Metrics for AI Impact

33:06 Challenges in Knowledge Management: Tackling Tacit Knowledge

35:56 Upcoming Events in 2025

37:26 Conclusion