@futureislands kicking off summer in amazing style.

via Instagram https://instagr.am/p/DLlDyKICTcQ/

I just posted this on Twitter as @ rvanbruggen

This was the text:

Crossposting this from #bsky: Achievement unlocked! I just published my first #blogpost on the #devrev blog. It's about how, in order to be truly productive, #ai #agents will need both #structured and #unstructured data. Read it over here https://t.co/RCED1CTk0I and let me know…

Crossposting this from #bsky: Achievement unlocked! I just published my first #blogpost on the #devrev blog. It's about how, in order to be truly productive, #ai #agents will need both #structured and #unstructured data. Read it over here https://t.co/RCED1CTk0I and let me know…

— Rik Van Bruggen (@rvanbruggen) Jun 30, 2025

from Twitter https://twitter.com/rvanbruggen

June 30, 2025 at 03:58PM
via IFTTT

I just posted this on Twitter as @ rvanbruggen

This was the text:

Crossposting this from #bsky: Amazing #learn and #lead event by #devrev in #amsterdam. First giving context and info, and soon we go into a #handson #workshop.. See https://t.co/spuwNkJ1yY for the original.

Crossposting this from #bsky: Amazing #learn and #lead event by #devrev in #amsterdam. First giving context and info, and soon we go into a #handson #workshop.. See https://t.co/spuwNkJ1yY for the original.

— Rik Van Bruggen (@rvanbruggen) Jun 25, 2025

from Twitter https://twitter.com/rvanbruggen

June 25, 2025 at 12:23PM
via IFTTT

I just posted this on Twitter as @ rvanbruggen

This was the text:

Crossposting this from #bsky: Here's a short article that I wrote about why I think #devrev 's layered approach to building #agentic #ai systems is so cool. See https://t.co/layCCWf8c4 or on LinkedIn... As always feedback welcome!. See https://t.co/LrpoTwcta2 for the original.

Crossposting this from #bsky: Here's a short article that I wrote about why I think #devrev 's layered approach to building #agentic #ai systems is so cool. See https://t.co/layCCWf8c4 or on LinkedIn... As always feedback welcome!. See https://t.co/LrpoTwcta2 for the original.

— Rik Van Bruggen (@rvanbruggen) Jun 16, 2025

from Twitter https://twitter.com/rvanbruggen

June 16, 2025 at 11:33AM
via IFTTT

Why Layers Matter in Agentic AI systems

Last week I published a quick post about the 5 layers of Agentic AI systems in the Enterprise. This article is going to be a more elaborate version of that post, because there is so much to discuss and explain.

AI agents are totally changing how businesses handle customer experiences, and the way they're built really matters. DevRev's gone a very specific and unique route for doing so, creating a multi-layered platform that's AI-first and agent-driven, all about making things better for customers. We're going to dig into how DevRev designed the system, look at each layer, and see why we made those choices, plus what it means for how the platform works and how people use it.

Why Layer in the first place?

Software layers create a structured architecture, making debugging, updating, and collaboration easier and faster. They isolate functions, reducing unintended side effects and streamlining maintenance. Without layers, the codebase becomes a tangled, chaotic mess, leading to difficult debugging, slow development, and jeopardized long-term viability. Just as organizing a workspace simplifies tasks, or layering a cake separates different flavours, software layers provide order and clarity, contrasting with the chaos of an unorganized system.

Layer 1: the foundational data

At its heart, the DevRev Agentic AI system is designed to be deeply intertwined with the lifeblood of your organization: its core enterprise data. We're not talking about a separate, isolated AI entity; rather, envision a symbiotic relationship where the AI actively draws from and contributes to the systems that power your day-to-day operations. Imagine it as a dynamic, bi-directional exchange where real-time updates from your product data, your support systems, your CRMs, and all your structured and unstructured data seamlessly flow into the AI, ensuring it operates with the most current and relevant information. By doing so, any insights or actions generated by the AI are intelligently fed back into those very systems, closing the loop and creating a truly intelligent and responsive ecosystem.

Layer 2: the 2-way sync mechanism

This two-way synchronization with existing, foundational Enterprise data sources - at DevRev we call this “Airdrop” - is absolutely critical. By grounding our AI in your existing data, we're starting from a position of strength. Your systems of record hold the validated, trusted data that forms the foundation of your business. We know that high-quality AI hinges on the quality of its data, and what better place to begin than with the information you already rely on? This approach eliminates the risk of AI operating on outdated or inaccurate assumptions, guaranteeing more reliable and insightful outcomes.


Furthermore, this strategy circumvents the need for disruptive, expensive, and time-consuming "rip and replace" projects. Forget about uprooting your existing infrastructure or forcing your team to adopt entirely new tools overnight. The DevRev Agentic AI system seamlessly integrates into your existing workflows, augmenting them rather than replacing them. This means a significantly smoother, more agile, and ultimately, faster implementation process. Your team can continue to utilize the software and platforms they're comfortable with, simply enhanced with the powerful capabilities of AI. This familiarity reduces the learning curve, minimizes disruption, and accelerates adoption, making the entire process easier and more efficient for everyone involved. Your domain experts remain in their domain, leveraging their expertise with enhanced AI assistance.

Layer 3: the knowledge graph

To transform foundational data into a dynamic and unified resource for the Agentic AI system, the various data streams from product data, support systems, CRMs, and structured/unstructured sources are consolidated into a connected Knowledge Graph. This knowledge graph acts as a central repository, interlinking the data points to create a comprehensive, real-time, 360-degree view of the client. This interconnection eliminates silos and ensures that the AI operates on the most current and relevant information.
Once synced, DevRev connects all the structured and unstructured data in a Knowledge Graph

Without this consolidated Knowledge Graph, the Agentic AI would be confined to individual data silos, resulting in fragmented insights and the risk of relying on stale or outdated data, hindering its effectiveness and limiting its ability to break down barriers between stakeholders. Building an advanced Agentic AI system demands a robust, interconnected knowledge infrastructure. The journey begins with the joining of diverse data streams – encompassing product information, support interactions, Customer Relationship Management (CRM) records, and a variety of structured and unstructured sources. These disparate data points are woven into a cohesive, dynamic Knowledge Graph.

A Knowledge Graph is essential for Agentic AI to function as an intelligent, integrated system. It provides a unified view of information, enabling deeper analyses and more effective outcomes. This centralized repository consolidates diverse data into a cohesive and updated source, offering a holistic client view. Without it, the AI relies on incomplete data, leading to superficial insights. A well-constructed Knowledge Graph allows the AI to leverage interconnected data for meaningful analyses and improved client relationship management. This unified approach unlocks the AI's ability to deliver insightful and accurate results.

Layer 4: shared platform services

Next in DevRev’s foundational architecture, is a layer that displays a strategy centering on shared platform services—specifically

• workflow orchestration,
• robust search capabilities, and
• advanced analytics

All these services are meticulously constructed and layered upon the unified knowledge graph described above. This strategic decision ensures that every application within the DevRev ecosystem benefits from a centralized repository of highly refined, thoroughly optimized, and consistently maintained tools. Consequently, developers are spared the burden of reinventing the wheel and can readily integrate these pre-built functionalities into their applications, resulting in streamlined development cycles, accelerated deployment timelines, and consistently high performance. This approach guarantees a uniform user experience across the entire suite of applications, fostering a sense of cohesion and intuitive usability.



Furthermore, the shared services model facilitates seamless inter-application communication and data exchange. By leveraging a common set of services that access and manipulate data within the shared knowledge graph, applications gain the ability to interact with and build upon one another's functionalities. This results in powerful cross-application synergies and leverages an environment where new, innovative features can be easily composed by combining existing capabilities. The interconnectedness fostered by this model dramatically enhances the overall value proposition of the DevRev platform, turning it into a truly integrated and collaborative environment.

Were it not for this strategic implementation of shared platform services, each individual application team would be forced to independently develop and maintain their own workflow, search, and analytics solutions. This would inevitably lead to a chaotic and fragmented development landscape. The resulting redundancy would not only represent a significant drain on engineering resources but also foster a proliferation of inconsistencies. Different applications would likely implement the same functionality in different ways, leading to discrepancies in behavior, conflicting data representations, and a disjointed user experience. Moreover, the lack of standardized integration points would severely hinder the ability of applications to communicate and share data, crippling potential integrations and severely limiting the platform’s collective capabilities. In essence, avoiding the shared services approach would not only incur a substantial cost in terms of wasted effort but also sacrifice the inherent advantages of a truly integrated and collaborative platform.

Layer 5: the end-user apps & agents

Right, so let's dive deeper into what makes DevRev's end-user applications a game-changer.

First off, BUILD is not just about making product management "smoother," it's about fundamentally transforming it. Picture this: instead of development and revenue teams operating in separate universes, BUILD acts as this central nervous system, connecting everything from initial feature requests to customer feedback and sales data. This seamless integration ensures that everyone is working off the same information, reducing misunderstandings and speeding up the entire development cycle. This isn't just about faster project timelines; it's about building the right product that truly resonates with customers because it's directly informed by real-world data and feedback loops.



Then you have SUPPORT, which goes way beyond just answering tickets. It's about creating a proactive support experience that anticipates customer needs. By having all customer data in one place and utilizing DevRev's AI capabilities, support teams can resolve issues much faster, often before the customer even knows there's a problem. This dramatically lowers support costs because issues are resolved efficiently, and it elevates the customer experience to a whole new level of satisfaction. Imagine, instead of endless back-and-forth emails, a support interaction that feels like a conversation with someone who truly understands your needs and has the tools to solve them instantly.

And GROW, well, it's about moving from guesswork to data-driven decision-making in sales and marketing. With access to a wealth of customer insights, market trends, and sales performance metrics, teams can craft personalized campaigns that truly resonate with their target audience. This isn't just about sending out mass emails and hoping for the best; it's about understanding who your customers are, what they need, and how to reach them effectively. By aligning sales and marketing efforts with real data, businesses can dramatically increase their lead generation, conversion rates, and overall revenue.

Summary

In summary, let's talk about what happens if you decide to skip these applications. You're essentially signing up for a return to the bad old days of siloed systems. All those disparate tools, all those data islands that don't talk to each other? They come back with a vengeance. Teams waste time manually transferring data between systems, information gets lost in translation, and critical insights are buried beneath layers of complexity. Plus, you'll miss out on leveraging DevRev's native AI agents, which can automate tasks, surface hidden patterns, and provide real-time recommendations. This isn't just a minor inconvenience; it's a major productivity killer. In today's fast-paced market, falling behind on productivity directly translates to a competitive disadvantage. And let's not forget about the impact on customer satisfaction. When systems are fragmented and support is inefficient, customers feel the pain. They become frustrated, their loyalty diminishes, and your Net Promoter Score takes a hit. In the end, bypassing DevRev's integrated suite means you're not just missing out on some cool features; you're missing out on a strategic opportunity to transform your business and stay ahead of the curve.

I hope this was a useful overview of my/our thinking around DevRev's architecture. As always, don't hesitate to reach out if you want to have a chat.

Cheers

Rik

I just posted this on Twitter as @ rvanbruggen

This was the text:

Crossposting this from #bsky: Over the weekend, I was reflecting on how customers build or buy #ai solutions. Should you #buildorbuy a technology these days? I thought I would write down some of my thoughts, and apply them to how #DevRev is using this approach today in the mark…

Crossposting this from #bsky: Over the weekend, I was reflecting on how customers build or buy #ai solutions. Should you #buildorbuy a technology these days? I thought I would write down some of my thoughts, and apply them to how #DevRev is using this approach today in the mark…

— Rik Van Bruggen (@rvanbruggen) Jun 2, 2025

from Twitter https://twitter.com/rvanbruggen

June 02, 2025 at 10:53AM
via IFTTT

The Enterprise Dilemma: Building vs. Buying AI-native CX Solutions

In today's every changing and evolving business landscape, enterprises face a critical decision when it comes to implementing AI-native CX solutions: should they build custom solutions from scratch, or buy existing platforms?

The Traditional Build Approach to AI in CX

Building custom CX solutions offers enterprises complete control over their implementation: the can fully customize their implementation, perfectly align with specific business processes, maintain proprietary intellectual property, and have direct control over feature development.

However, this approach comes with significant drawbacks: building custom AI solutions comes with significant challenges including high development and maintenance costs, extended time-to-market, and resource-intensive updates and improvements. Plus: whether you like it or not, there’s quite a bit of complexity to creating a solid AI solution - which is only to be met with significant skill!

The Traditional Buy Approach to AI in CX

Purchasing existing CX solutions provides several immediate benefits. Companies can deploy these solutions rapidly, leveraging proven functionality that has already been tested in the market and that has been engineered by highly specialized staff with very specific skills. These solutions come with regular updates and improvements managed by the vendor, and typically require a lower initial investment compared to building from scratch.

However, this approach also comes with notable limitations. Organizations often find themselves restricted by limited customization options and become dependent on the vendor's roadmap for new features and improvements. Additionally, there's a risk of misalignment between the pre-built solution and specific business needs, which can impact operational efficiency.

DevRev’s Hybrid Solution: A New Paradigm, validated by the industry

Modern platforms like DevRev are pioneering a hybrid approach that combines the best of both worlds: lots of out-of-the-box functionality that relieves you of the boring infrastructure related tasks, combined with extensive customization capabilities to tune the platform to your needs.

This innovative approach offers several distinct benefits: core functionality is available immediately while maintaining the flexibility to customize and extend the platform according to specific business requirements. We can summarize this like this:

This is not just DevRev saying this: McKinsey's 2024 "State of AI" survey shows that 75% of enterprises prefer solutions that offer both out-of-the-box functionality and extensive customization capabilities. This confirms the trend that we have seen: it aligns perfectly with the hybrid approach offered by modern platforms like DevRev.

Conclusion

The traditional build-vs-buy dichotomy is becoming obsolete. Modern enterprises need AI-Native solutions that combine immediate functionality with the flexibility to adapt to specific business needs. Platforms that offer this hybrid approach, like DevRev, represent the future of enterprise CX solutions.

By choosing a hybrid solution, enterprises can accelerate their digital transformation while maintaining the ability to differentiate their customer experience - truly offering the best of both worlds.

Let me know if you have any questions or comments. Would love to discuss.

All the best

Rik