Completely agree that as bot builders we have a lo...
# 💼business-advice
Completely agree that as bot builders we have a lot of room to grow into a largely untapped global market and leverage pre-existing niche skills/strengths through our networks. 🌐 

 One of the reasons I prefer Botpress over other competitors, and why I’m betting on bot-building platforms, is code. I feel that the large selling point of platforms is standardizing the backends and simplifying the conceptual UI/UX for clients with the visual builder. As to the differentiator for Botpress, I like that there is a lean-in attitude when it comes to code, versus some other platforms (like Voiceflow), which to me, feel like they’re abstracting away too much programming artifacts. Every year new developers join the industry; and the trends I see are moving to low-code and platforms which offer a more stable backend for developers and clients. That being said, code and programming still underpin the nuts and bolts of any software application (AI or otherwise) and no matter how much abstraction there is, what’s under the hood is code. I think making programming intuitive and accessible is an advantageous strategy. I’m optimistic that platforms are not just a bridge to the future, but the a shape the new paradigm will take. There are so many examples: website, e-commerce, or blogging; these kind of platforms are still popular and in use today for good reason. Echoing this vibe, Andrew Ng of DeepLearningAI recently said this in The Batch newsletter: > “I think the complexity of Python package management holds down AI application development more than is widely appreciated…. This is annoying but manageable for experienced developers, but creates a lot of friction for new AI developers entering our field without a background in computer science or software engineering.”
 Ng ultimately doesn’t have a resolution for his observation, but his words resonated with me and felt applicable to this thread.

 At the end of the day, whether a Bot Builder #OneTeam is a code-pro or low-code developer, providing the human point-of-contact to the non-technical world is invaluable service (and something the big tech companies hire people to do too)! Understanding how to manage the trade-offs of how and when to best deploy AI (vs NLU for example) in conversation design will be among the kinds of skills and perspectives that can contribute to efficiency at scale for clients; and ultimately to sustainably provide the best results to end users. 🤖 Additionally, providing on-going support and maintenance is perhaps the hidden treasure of bot building. Software updates are already kind of a hassle for a non-technical operator… and AI software updates are even more fickle and frequent! 🪲 To close on a fun note, the killer feature I’m looking most forward to seeing rolled out from the Botpress team is ability to choose from different (foundational and/or fine-tuned) models! Sure, OpenAI is a leader in the state-of-the-art, but a lot of superb SOTA models have been released in the past year. Offering ways to avoid model-vendor-lock-in will allow for competitive advantages to builders & clients in many ways. (Of course I’m looking forward to multi-model and multi-modal features too! 🙏🏼🤩 )