Prompt-over-Code Design Method

Agentic We Are – And our Agent Development Method can tell.

The Prompt-over-Code Design Method (PoC) explains how Agentic AI solutions intelligently integrate into each piece of a process or workflow. Traditional coding approaches make designing non-linear processes—such as innovation, transformation, or complex banking solutions—inherently tricky.

How did we get there?

BlueCallom has developed early versions of Agents since 2021. Back then, we called them “Multifunctional Prompts,” for lack of better words. We were using GPT 3, which was over a year before ChatGPT was released. When ChatGPT was announced and the new GPT3.5 and its new API structure were released, we understood that this technology would create a seismic shift over the next five years. It would allow us to build highly sophisticated business solutions that no company could build with conventional software.

After working daily with our “Multifunctional Prompts” for over two years, we realized that our way of thinking has changed. We are less concerned about coding and more about flexible and intelligent processes. We learned to give the GPT/LLM system more freedom, and in return, we got more intelligent responses. In early 2024, the talks about AGENTS became more critical. Even though our “multifunctional prompts were kind of Agents from the beginning, there was more to those new Agents. We considered building Agents with Python, like roughly 90% of the agent-developing community. But wasn’t that a huge step back to narrow, stiff, and rigid coding?
The hardest decision in our software career

Shall we write Python code that sends prompts to the LLP whenever needed, or should we prefer prompt control over the workflow and only use code when needed? Of course, both prompts and code are needed. But, in the long term, would we be better off with prompt-based intelligent processes or go back to hard-coded processes where we came from? At the time, it felt like a brutally hard decision. And there is no meaningful middle way.

All Or Nothing.

We made a conscious decision to abandon traditional coding and embrace prompt-over-code (PoC) as our new paradigm.  Please don’t confuse Prompt-over-code with “No-Code”. It represents a core design principle that leverages prompts and large language models (LLMs) to drive intelligent workflows. Instead of relying solely on rigid code structures, we use prompts to manage processes and call specific functions only when needed. These functions may include algorithms, deep learning mechanisms, API access, or natural language processing tasks. This approach allows us to build dynamic and adaptable workflows, which was impossible with conventional software before.

 

Prompt-based process design looks more like a network

A quantum Leap with the Prompt-over-Code  Method (PoC)

Intelligence

Unlike conventional software, which has used AI for algorithms and machine learning in the past, PROCESSES, HUMAN INTERACTIONS, DATA QUALITY, etc., agentic AI is a powerful instrument for building non-software Intelligent Processes, Intelligent Human Interaction, and Intelligent Data Handling.

In that context, intelligence means autonomous processes, processes that can learn or adjust based on external changes, processes that humans can quickly adjust without coding, and processes that may change based on better knowledge of those processes elsewhere in the organization.

Agentic AI Solution Design with PoC brings the INTELLIGENCE in the forefront of process design.

Intelligent human interaction means typos and minor mistakes can be handled without responding. Human entries no longer need to follow exact date entry rules or repeatable data to satisfy the nonintelligent software process. Humans don’t must comply with the rules of the machine but with a common sense that is given to the handling of the AI.

Data Quality is an epic topic in stiff software code. All too often, a telephone number must be entered in a specific format, the order must be accepted according to certain rules, and filling out a form requires a degree. Intelligent Data Quality management understands what the data ‘means’, not only what it states.

Complexity

The level of complexity in modern enterprises has outgrown human comprehension. Some structures, like supply chain processes, are over 30 years old. However, changing those processes may require building a new logistics organization. R&D Departments may be perfectly organized and managed, and their output rated with fine-tuned KPIs. Yet modern innovation processes could not be used because the internal changes would be too massive. Agentic AI may introduce new processes entirely independent of existing and old structures and can be managed without integration.

We can’t remove complexity with the same tools that created it. Therefor we decided for the PoC method.

For decades, enterprises have known that their data structures must be cleaned up. Processes increasingly show friction, and productivity is slowly but steadily declining. Yet, no enterprise with more than 250 employees is known for not trying to clean up internal data and processes. The pain continuously grows—to the degree that innovation and other key growth drivers are pushed into the background.

Unlike conventional AI, which uses hard-coded algorithms and enormous numbers of iterations, the power of Agentic AI comes from its intelligent ways of assessing, suggesting, reasoning, and creating solutions in one process, even autonomously.

Sharing our secret sauce

After a year, we now experience another significant opportunity: Development partners can far more easily build connections to our agents and solutions when all processes are led by a prompt-based protocol than when connecting one set of code with another. Compatibility, Testing, Data Exchange, and orchestrating a vast multiagent environment is far easier with PoC than with hard-coded processes. We share our “secret source” today because our partners will need to know how to build agents to be connected to existing agent networks and digital AI asset management concepts.

Prompt-over-Code Method with ChatCPT

It will understand it and use it to realize your ideas. Moreover, it quickly tests the code on its own platform, so you know it works when you integrate it into GPTBlue on BlueCallom.

Tell ChatGPT to use the Prompt-over-code method

WHAT YOU NEED TO KNOW
ChatGPT can help build brilliant AGENTS using the Prompt-over-Code method. Once you share your needs and ask it to help create the solution, it develops all necessary prompts for the primary process, as well as the necessary Python code in case code is needed. You can even test the result on ChatGPT.

HOW YOU TELL CHATGPT
Obviously, ChatGPT needs to know about this method, as it is not yet part of its model. So we had a lengthy discussion about what it should do, and it did a perfect job. It was so cool that we were congratulated on that method. We asked how others could explain the method to ChatGPT efficiently. It immediately wrote a description of what you should tell it. Click here for the description.