Beyond the Hype: How DerbySoft is Building AI for the Realities of Travel Commerce

Travel technology is in the middle of a gold rush. Nearly every platform, startup, and legacy provider in the hotel distribution ecosystem has declared itself “AI-powered.” Funding rounds are being secured solely on the strength of AI narratives. The energy is real, but so is the risk. Because in the rush to claim the future, too many companies are skipping the hard, unglamorous work that makes AI useful.

At DerbySoft, we have made a deliberate choice to take a different path. Not a slower one for the sake of caution, but a more grounded one, built on a simple conviction: AI that is not anchored to accurate, structured data will eventually fail the people who depend on it.

I want to share where we are on that journey, what we are building, and why we believe our thoughtful vision will matter more than speed in the long run.

The Boring

The least exciting part of AI is also the most important: data quality.

Hotel distribution has a persistent problem that predates the current wave of AI enthusiasm. Property data across systems, languages, and channels is wildly inconsistent. Room types are labeled differently across platforms. Amenities are described in dozens of ways. Policies vary in format and granularity. For travelers, this creates confusion. For hotels and distribution partners, it erodes trust and conversion.

Our AI-powered Content Solutions platform is built around three capabilities that address this directly. AI Fill-In automates data entry and content updates within the management system, keeping property information current with minimal manual effort. AI Extract transforms long, unstructured property descriptions into rich, structured data, converting free-text paragraphs into clear, searchable attributes like precise room sizes, accurate bed types, and detailed amenity lists. And AI Review scores content for completeness and relevance, giving hotels and distribution partners clear, actionable insight into gaps so they can improve quality before it reaches the traveler.

Together, these tools connect hotels and distributors through a single, trusted content platform. Hotels maintain central control over their content. Distributors access accurate, enriched information at scale. The result is consistency and confidence across every channel.

None of this makes for a dramatic press release, but it is the layer everything else depends on. Deloitte’s research on generative AI in the enterprise reinforces this point: data quality remains one of the most significant barriers to scaling AI effectively. We took that finding seriously and built accordingly.

Protecting Revenue Integrity

Rate parity is another area where we are embedding intelligence directly into infrastructure. Rate discrepancies across channels remain a persistent concern for hotels and their distribution partners. When pricing breaks down, trust breaks down with it, and margins suffer.

We are developing AI-enhanced rate monitoring and parity validation tools designed to detect anomalies faster, reduce the burden of manual auditing, and protect pricing integrity. This is not a standalone AI product. It is intelligence woven into the distribution workflow itself. The objective is not to showcase technology. It is to eliminate inefficiencies that directly affect revenue performance.

Automation Grounded in Partnership

We are also applying AI to automate content workflows and product updates, and our recent collaboration with Agoda is one example of this in practice. While we are not yet ready to share detailed performance data, internal work confirms that AI has meaningfully reduced manual intervention in product update processes.

This reflects a principle we hold closely: demonstrate results first, amplify later. As more performance data matures, the opportunity for deeper case-based storytelling will follow. But we are not going to get ahead of what the evidence supports.

Beyond Price: The Shift Toward Experiences

Perhaps the most significant evolution in our AI thinking is the move toward behavior-driven personalization.

For years, travel shopping has been dominated by price-first sorting. The lowest rate wins the click. But traveler expectations are shifting. Booking confidence increasingly depends on richer context: loyalty program benefits, breakfast inclusion, flexible cancellation, local experiences, and room attributes that go beyond square footage.

Our content platform already supports the inclusion of these richer attributes. AI layers can then help match them to individual shopper behavior, surfacing the information that actually drives decisions rather than defaulting to price alone.

McKinsey research suggests that personalization can drive revenue uplift of 5% to 15% and improve marketing spend efficiency by 10% to 30%. In travel, that uplift depends heavily on structured content and intelligent filtering, which is precisely the foundation we have been building.

This is also why I participated on an industry panel in Milan focused on rate choice and richer content in the booking experience. The message was straightforward: booking confidence improves when travelers see complete, robust, trustworthy information, not just the lowest price.

AI Voice and Conversational Commerce

Some of the most labor-intensive friction in travel happens after the booking is made. Confirming that a property has the correct reservation and payment details before a traveler arrives at check-in. Retrieving folios when APIs are unavailable or when invoices need correcting. Handling urgent booking modifications when automation fails or when a traveler’s plans change at the last minute. These are high-volume, time-sensitive tasks that have traditionally required human agents making manual calls, often at significant cost.

Our AI Voice Agent is designed to handle exactly this kind of operational work. It verifies booking and virtual credit card details directly with properties. It collects invoices. It processes modifications. And it adapts to each partner’s specific workflows, scaling to support millions of bookings per year without requiring proportional increases in headcount.

Companies using the solution in pilot programs are seeing 70% to 90% reductions in costs associated with manual calls. In pilots with leading travel management companies, over 75% of bookings were completed through the AI Voice Agent, eliminating the need for manual intervention entirely. Those are not projections. They are operational outcomes.

The downstream effects matter just as much. When booking verification and invoice collection happen reliably and at scale, the entire traveler experience improves, from check-in through checkout and reconciliation. Payment success rates increase. Collection processes tighten. And the people who were previously spending their days on routine calls can redirect their time toward work that requires human judgment.

A Platform, Not a Patchwork

One theme that comes up repeatedly in our internal strategy discussions is coherence. The market does not want a proliferation of disconnected AI tools. It wants a unified system where content intelligence, rate integrity, normalization, behavioral personalization, and conversational interfaces all operate within the same ecosystem.

That is what we are building. And it matters because one of the most cited risks in enterprise AI adoption is fragmentation. McKinsey/Skift research highlights integration complexity and data silos as primary obstacles to scaling AI reliably. By building each new capability on top of an existing structured data layer, we reduce the risk of deploying tools that lack dependable backend support.

Earning Trust in an Era of Overpromise

The current AI landscape rewards bold claims. Companies that aggressively pitch transformation stories attract attention and funding. But IBM’s Institute for Business Value reports that only 25% of AI initiatives achieve expected ROI at scale. That gap between promise and performance is where trust gets lost.

We have seen this pattern before in technology cycles. The companies that endure are rarely the loudest. They are the ones that build carefully, validate rigorously, and scale only when the foundation can support it.

Our trajectory is designed around that belief. We want to build trust with our clients by delivering measurable results. We want to avoid costly rework caused by rushed deployments. We want to make sure our data foundations are strong before we scale automation further. And we want AI to integrate into the broader travel operating model, not sit alongside it as a disconnected novelty.

Many AI failures stem not from algorithm limitations but from weak data governance and fragmented system integration. 

The Long Game

AI innovation in travel will not be won by speed alone. It will be won by durability.

Our roadmap reflects a philosophy that I believe our industry increasingly values: innovate but validate. Automate, but integrate. Personalize, but ground it in accurate data. And of course, the end result is ROI.

We are not trying to be the loudest AI voice in travel. We are working to be one of the most structurally prepared, because when the dust settles, that is what will matter most.