It might be hard to believe, but it hasn't even been two years since ChatGPT's first general availability release.
A lot has changed since then.
ChatGPT brought large language models into the mainstream, ushering in a wave of change into the world of software. One of the many new opportunities LLMs unlocked is the concept of Service as Software.
Whereas Software as a Service (SaaS) was the prevailing way to deliver software in the '00s and '10s, Service as Software (SaS) stands to pave the way in the late '20s.
Think of traditional B2B SaaS applications like Salesforce, Canva or Quickbooks. These applications provide services like customer relationship management, photo editing or bookkeeping.
They can make you more productive, but at the end of the day, you still have to do the work.
Service as Software turns the old-fashioned SaaS model on its head. Instead of requiring a human to spend time and effort doing menial tasks to maintain a CRM or accounting software, the SaS uses a generative AI agent to perform those functions as if a person was doing it.
This, in turn, frees up professionals to focus on more engaging, meaningful work.
There is a lot of media focus on the LLM arms race between OpenAI, Google, Meta and Amazon, as each of the tech giants has tried to get their hands on as much training data as possible.
OpenAI's GPT-4 model reportedly has 1.7 trillion parameters, 1,000x the size of GPT-3. When it's eventually released, GPT-5 will surely have a whole lot more.
There's no real opportunity in trying to compete with the behemoths. Instead, it's all about using the models they've already developed to power your SaS applications.
Here are a few examples of Service as Software applications making waves:
Archie by 8base helps software teams flesh out fully formed product and engineering requirements based on simple, plain language inputs.
One of the most common pitfalls in software development is ambiguous requirements. This can slow down the development process or even set it back if it results in bugs or defects.
Archie uses multiple LLMs to deliver comprehensive requirements every time, helping teams see around corners and even boost their creativity.
Archie is free to try. Get started here.
One of the most challenging aspects of sales is having the right answers at the right time.
Docket estimates that sellers spend more than 30% of their time wading through old documents to find answers to questions. This is inefficient, and it can lead to lost sales if AEs don't have the right answers at the right time.
Docket easily integrates with more than 100 different sources to give AE's a single source that they can easily ask questions to, as if they were pinging a coworker in Slack.
Dropzone estimates that less than 10% of cybersecurity alerts are thoroughly vetted.
Alert fatigue is real. Data is fragmented, analysis is tedious and there's a seemingly never-ending stream of alerts.
Dropzone provides a turnkey solution that does end-to-end analysis of all cybersecurity alerts, without any prompt engineering or fine tuning needed.
Healthcare is a well-known morass of paperwork.
Tennr is a HIPAA-compliant SaS that uses healthcare-tuned LLMs to parse documents, respond to them when needed and route them to the correct recipient.
This can streamline everything from patient onboarding to insurance claims, referrals and prescriptions.
Of course, the place to start is with an idea. What's a common challenge that people face, which can be solved by a generative AI agent acting as a human?
Your idea doesn't need to be huge or world-changing. In fact, it's often better if it isn't. It just needs to solve a specific problem for a group of people better than anyone else can do it.
As you work through the ideation process, Archie can be a valuable asset. It can not only help you think through your idea and its viability, but Archie can also provide detailed technical documentation and UX wireframes to help you begin building your solution.
Then, when it's time to start building, the technology you use will be dependent on the solution you are building.
8base's low code backend-as-a-service as a great option to get up and running quickly. It easily integrates with LLM APIs like OpenAI's GPT-4, Anthropic's Claude and Google's Gemini, and you can seamlessly build with your frontend of choice.
Service as Software is transforming the traditional SaaS model by using generative AI agents to perform tasks traditionally done by humans.
This frees professionals to focus on more meaningful work and offers a promising new direction for software development.
We're excited about helping you achieve amazing results.
