The Model is the Product: Service as Software (SaS) and Why It Matters
Service as Software (SaS) redefines digital experiences, shifting focus from tools to outcomes based on Domain-Adapted LLMs.
Every year, new trends pop up, and you’re left wondering: should we jump on the bandwagon or sit this one out? But I believe this topic is different.
Service as Software (SaS) isn’t just another catchy acronym to slap on a PowerPoint slide—though, yes, it deserves a place on those slides. It’s a fundamental shift in how we design and manage journeys and the experiences that result.
This is one bandwagon your organisatoin wants to be on. The investments your organization has made on Artificial Intelligence—people, infrastructure, knowledge, compliance and launched experiments—(hopefully, they’re in place) are building the foundation as the Service as Software starts to unfolds.
The SaS concept goes beyond tools or platforms; it redefines what digital services can be, opening up an unprecedented opportunity for creativity and growth.
SaS isn’t just a trend—it’s a transformation in how we need to think about building, and scaling digital touch points.
How software morph into service providers?
The SaS Paradigm: Moving Beyond Software as a Service towards a System of Agents.
For decades, enterprise software has focused on structuring and storing data rather than understanding and assisting people. Employees find themselves bogged down by data entry, the loss of crucial business context, and systems that only skim the surface of business intelligence.
The SaaS model as we know (and use it) falls short of helping businesses to adapt and evolve; it’s more about operational support than innovation and service transformation.
The Systems of Agents: (almost) autonomous AI-driven entities (AI-agents) that do more than just assisting based on provided input. They take action. These agents parse emails, understand calls, process documents, and turn unstructured information into structured insights that can actually drive business and improve customer service.
AI-agents are the next evolution of enterprise tools – not just supporting workflows but redefining them in direct collaboration with their customers.
A $4.6 Trillion Opportunity
We’re looking at a market set to reach $4.6 trillion, driven by AI-powered agents with domain-specific expertise as Foundation Capital researched in Q1/2024. Unlike traditional SaaS products, these systems don’t merely support users; they work in tandem, autonomously handling tasks and delivering end-to-end outcomes.
For instance, a System of Agents might include specialized AI agents like SDRs (Sales Development Representatives) qualifying leads, SEs (Sales Engineers) handling technical assessments, and AEs (Account Executives) closing deals. The agents continuously learn from each interaction and each other, creating a cycle of improvement and efficiency that is unprecedented.
Service as Software (SaS): A New Kind of Product Experience
Imagine this: instead of using a travel booking website, you have a personal AI travel agent who curates, plans, and manages every detail of your trip. This is what Service as Software offers.
Rather than purchasing software tools, you’re subscribing to services delivered by agents trained to act on your behalf.
In this new paradigm, SaS companies won’t be selling features – they’ll be selling outcomes. Each agent will be empowered by advanced Large Language Models (LLMs) and Multimodal Models (MMMs) that adapt to customer needs in real time. Instead of static, one-size-fits-all LLMs, these agents will be fine-tuned with domain-specific knowledge to handle the nuances of complex service scenarios.
Why Domain-Adapted LLMs Are the Game-Changers
Generic LLMs are impressive, but they’re not yet ready for specialized tasks. Without customization, these models act as NPCs (non-playable characters) in customer interactions, failing to grasp the specific contexts and subtleties that make services valuable. Domain-adapted LLMs, however, are fine-tuned to specific roles, understanding the unique demands and languages of fields like sales, customer support, or legal consulting.
Think of it like cooking: data is the main ingredient, analytics is the seasoning, and fine-tuned LLMs are the final dish. This approach ensures that AI agents aren’t just task-completers but adaptive service providers, deeply knowledgeable in their respective areas.
As I work with a leading global real estate brokerage, I see how SaS can go even further. Imagine an AI-agent designed specifically for real estate transactions. This domain-adapted LLM isn’t a generic chatbot; it’s fine-tuned with in-depth knowledge of real estate regulations, market trends, local specifics, and property transaction processes. Trained on real estate transactional data, contracts, listings, and industry-specific jargon (better exclude selling sunset episodes from training the LLM), it fully understands the nuances of the field, enabling it to support customers in the best possible way. Finally with the goal to hand over an almost transaction ready customer to a real agent to take the last mile and earning their commission, with full transparency on the AI-agents performance from the first customer interaction on.
Why The Model is the Product?
With Service as Software, customers no longer buy software; they subscribe to agents (the service) with specialized knowledge, skills, and adaptability.
Each agent embodies a blend of data science, industry insight, and operational intelligence. The LLMs behind these agents are continually refined to deliver the exact expertise customers need, rather than a generic solution.
The implications of the Service as Software (SaS) transformation are significant.
Instead of purchasing SaaS products with static features and functionalities, businesses will rely on agents capable of continuously improving and customizing their services in real time. This approach is not only more efficient but also far more responsive to evolving customer needs.
The mindset shift we as design and digital leaders need to tackle is that it’s no longer about the next feature or tool we design and build. As these won’t deliver the level of value businesses and customers will expect.
Our focus should be on the tangible outcomes delivered by a system of AI agents empowered by domain-specific, trained LLMs.
A model trained with your company’s knowledge and expertise will become the product that drives business growth and redefines customer experience. A model that is so unique to its capabilities, to make a clear differentiation to the competition with the ultimate goal in outperforming them.
While this may sound unusual at first, it will soon become the new normal as artificial intelligence moves from bots to agents, delivering real, measurable service value to your customers.
