Python Automation

The rise of Python in the era of AI Automation

Every market needs time to reach maturity and move to the next steps. The Intelligent Automation market is similar to that. When companies started automating their process, by that time, there was a promise that every business user could automate everything with no need for developers.

After some years, companies evolved their automation projects, moving from individual tasks to function-specific processes and finally reaching processes across internal and external ecosystems, where the most significant challenges and the best outcomes reside, as you can see below in this chart adapted by us from Gartner:

During this evolution, there is one truth that companies discovered along the way:

Complex automations demand coding skills, even on low-code platforms.

Business users can build simple automation, but coding skills are needed as soon as we reach some level of complexity. Another evidence is that the most prominent players in this market rely on a vast ecosystem of partners for implementation (UiPath, for instance, has 6,000 partners). Why is this ecosystem so extensive if low-code was designed for business users (a.k.a citizen developers)?

The answer is that complex automations are software projects, and software projects demand coding skills. So, companies need these skills, even using low-code platforms (after all, low-code is just an abstraction from a code, in other words, just someone else’s code). So, some of the options here for complex automations are:

  • Hire developers to work with low-code (it works, but companies face challenges on turnover)
  • Hire consultancy companies + internal developers.
  • Build internal capacity for coding skills.

No matter the option, at the end of the day, your company already has coding skills somewhere to deploy these projects.

The question here is, if you already have coding skills, why are you not coding? The answer: Governance

In 2018, we discovered that companies were not using code as the main stack for automation due to the lack of a governance and powerful orchestration solution that could bridge the gap between technical teams (responsible for building and maintaining these automations) and business users (key stakeholders who need to have autonomy and control over their automation).

That’s precisely what we solved with BotCity. We built an ecosystem of products to deliver two things: Productivity for developers (with our developer tools) and Enterprise Governance (business users connected to technical teams).

We focused on code-based automation (Python, Java, JS, C#), but why do we talk so much about Python?

  • Versatile: Python is the glue for the whole ecosystem of Intelligent Automation. You can rely on more than 450.000 Python libraries globally for anything you want. Sometimes, we hear, “Is this operation that I made in the XYZ platform possible to do in Python?” We always answer that the question should be flipped, asking if something we can do in Python is possible in low-code platforms.
  • Learning curve: The Python learning curve is fast, and with BotCity frameworks, plugins, academy, and all the learning resources now with copilots, the curve has flattened.
  • Open the Path for AI: If your company wants to be AI first, Python needs to be in your daily operations. This skill in your team can open the path for Machine Learning, Gen AI, NLP, and computer vision, and then we are talking about Intelligent Automations.
  • User base: If you go on Linkedin for the top 4 RPA platforms, you’ll find 200.000 users globally. It would be best if you found users skilled in a specific platform. There are 15 MILLION Python developers; the knowledge is Python, not specific to any platform.

The Gen AI boom – shaping the automation future

Everything accelerated with the Gen AI explosion. From the new possibilities towards complex unstructured data to the easiness of code. Let’s talk a little bit about these topics:

  1. Gen AI accelerated Python adoption

Why wait for a vendor to implement a solution in an ever-changing landscape? Coding in Python, you can go straight to the source and get access to SDKs and open source libraries, not just on Gen AI but also on machine learning, computer vision, and NLP.

We saw a huge increase in companies building capacity in Python, such as our customers Bayer and LG. Bayer trained hundreds of employees, and LG trained students who will work as Digital Transformation agents.

  1. Coding is easier than ever. Governance is more complex than ever. The shadow IT rises

Anyone can ask ChatGPT or Gemini to generate Python code. This led to an explosion of business people learning to code now that coding is easier than ever. On the other hand, there’s a huge shadow IT wave rising (we already have customers who adopted BotCity just to eliminate Shadow IT from Python automation), and governance will be critical.

  1. Intelligent automation for real

The new capabilities acquired by Gen AI are enabling the possibilities to build even more complex automation that demands steps like:

  • Data conciliation;
  • Decision-making in processes using Machine Learning;
  • Data and document classification;
  • Inform and explain using natural language processing;
  • Navigate legacy applications using computer vision;

The outcomes will be even more significant in the dawn of this new era, but in the end, the complexity will be a developer’s job.

  1. AI-Agents will eat low-code in the long run

AI agents, our APA (Agent Process Automations) will rise with the same promise as low-code: “No developers needed, we can solve everything.” Again, the same maturity cycle will happen, with the overpromise undelivered in complex scenarios. AI agents will perform well in low-complexity situations, from user task situations to medium-complex processes in a department.

How about the complex process? These processes will remain developers’ jobs, but now, with new copilots that will enhance productivity and help build faster than ever.

  1. Low-code and Python can coexist as the best toolbox

Finally, is important to say that what matters to reach results is to have the best toolbox. The best tool for the best use case. What we are seeing globally is a multi-vendor approach, as we as BotCity coexist in several customers with UiPath, Automation Anywhere, Blueprism and Power Automate. This way you can have the best of two worlds: empowering developers and business users for the correct use cases regarding complexity, making your TCO more efficient, reducing the vendor lock-in and the governance and scalability needed to accomplish your automation goals.

Want to know more about Python governance and orchestration? Go to botcity.dev to know our solutions.

Leave a Reply

Discover more from Blog BotCity - Content for Automation and Governance

Subscribe now to keep reading and get access to the full archive.

Continue reading