From Code to Conversation: How GPT is Changing the Game on Power Platform
Vidya Sagar Alti
Technology Head - Power Platform, AI.Cloud Unit at Tata Consultancy Services
The evolution of code abstraction has been significant in recent years, and large language models like GPT is taking it to the next level now. The shift from custom to low code has made app development more accessible to a broader audience. However, with the introduction of natural language authoring, humans and machines can now work together to build applications more efficiently than ever before.
Taking center stage is the Microsoft Power Platform – the new features leveraging GPT takes app development to a whole new level, like we have never seen before. The latest in the chain of advances for Power Apps, Power Automate, and Power Virtual Agents is Copilot.
Copilot in Power Apps:
With GPT-Power Platform joint workings in 2021, Power Apps Ideas for advanced PowerFx formulae integration with natural language and Power Apps Express Design for generating app screens using image and Figma design files were introduced - for creating arithmetically sound & rich UI applications.?
Copilot takes this to a new high by leveraging Power Apps and GPT to make cutting edge developers out of citizens. Copilot builds an app, including the data behind it, through multiple steps of natural language instructions. A simple command like "Create a site inspection app" is enough for Copilot to create Dataverse tables and generate a fully functional & code compliant app.
And take it a notch up with the integration of AI-powered Power Virtual Agents in Power Apps. With the embedding of AI chatbots to Power Apps-based applications users can discover conversational insights on a click of a button.
Copilot in Power Automate:
Copilot allows users to build Power Automate flows using natural language - providing a dynamic and ‘lets-talk-and-test’ conversational model of workflow building. And Power Automate Desktop or the RPA modernization tool now allows users to effortlessly summarize lengthy documents and extract pertinent information with the "Create text with GPT" functionality.
Copilot in Power Virtual Agents:
Copilot in Power Virtual Agents enables topic creation using natural language processing, including suggestions on relevant trigger phrases, questions, entities, messages, variables, and other logic models. This brings in a new era of bot creation allowing users to program and provide ‘close-to-human’ responses and resolutions. Additionally, the "Boosted conversations" feature helps bots find and present information from external data sources, even if that topic hasn't been formally created – like a safety net to ensure no questions return an empty response, while capturing new data sets at the backend for maker suggestions.
Azure OpenAI Service in AI Builder:
领英推荐
The Azure OpenAI Service feature such as text summarization makes it effortless to run through lengthy documents, extract knowledge snippets, and create a succinct brief for effective content absorption. Which of course can be used for any chat-based query operations too.
In summary, the low-code development, combined with embedded generative AI, is revolutionizing the tech industry, providing a new level of accessibility to individuals and IT departments alike. With Copilot and other similar platforms, it seems very possible for everyone to address their unique objectives by creating powerful, intuitive, engaging, and innovative solutions.??
If you are still thinking about making low-code mainstream in your organization, its time you put a stop to the apprehensions and get into the game. Leading industry analysts expect the low code market to grow at a compound annual growth rate of 28% over the next five years, which seems very real given the advancements we are seeing.
In conclusion, low code does not seem to be a choice but a mandate to all. If agility, tech-savviness, sustenance, differentiation, innovation, growth are your go-to-words – then low code, AI, GPT are inevitable. But of course, this is my take on this and I am just as keen to hear your thoughts. In your opinion is low code & AI the future of our workplace?
References: