In today's rapidly changing business landscape, success comes to those who stay ahead of the curve. This means maintaining a laser-sharp focus on marketing, brand and reputation building, and deeper, more comprehensive market penetration for any pharmaceutical manufacturing firm.
On this front, generative AI has become a boon for marketing teams. It saves them significant time and effort in creating content and go-to-market materials for their campaigns, such as blogs, graphics, brochures, videos, and more. This helps marketing professionals focus on areas that require their undivided attention.
A pharmaceutical major partnered with Birlasoft to engage Generative AI in creating first drafts of hyper-personalized visual and text-based content for promotional materials, such as videos, graphics, and social media posts, based on the campaign objectives and target audience. Appropriate prompt engineering guardrails ensured that the content produced was within policy and compliance guidelines, ensuring messages and promotional materials were truthful, balanced, and accurately communicated.
The Challenge
Understanding the content generation landscape of the pharmaceutical manufacturing major
- As with most global organizations, the customer had multiple pharma commercial operations teams to write blogs and publish campaign content to meet the high demand for the rapid creation of novel content for campaigns and other marketing activities.
- Creating unique content was usually time-consuming and prone to errors. It also often lacked personalization, which rendered it less effective and persuasive.
- The organization leveraged several SaaS tools and products to generate, curate, and publish campaign drafts. The sprawl of the tools added greater complexity to the process, making it challenging to meet the demand. Additionally, this meant that the organization dealt with multiple overheads from multiple SaaS tools, making the processes inefficient in cost and effort.
- There was a need for an in-house platform that could help generate the first drafts of most of the content being produced by the team (e.g., brochures, videos, images, etc.). Creating personalized first drafts could cut down the timelines and enable faster creation of unique and compelling content.
The Solution
Building a generative AI powerhouse of creativity
Birlasoft's Generative AI program with the pharmaceutical major involved incubating a 'Generative AI-derived first draft' approach. This was aimed at reducing manual effort and dependencies on content-creation SaaS tools.
- Rapid PoCs were delivered using several frozen and open-source diffusion models and customized chain-of-thought frameworks.
- It started with PoCs for image generation, followed by PoCs for video synthesis, and talking avatars as creative assistants.
- Generation of product descriptions and detailed responses to queries
- The MVP was rolled out with an image and video generation application trained on advertorial use cases.
- An MVP for blogs and brochures drafting was created using a solid model i/o backend and LLM agent mechanism.
- The goal was to meet up to 10-20 percent of novel and contextual content creation through this platform.
- The tools and the stack that powered the transformation included.
- AWS Bedrock, Stable Diffusion SDXL Models, Huggingface, Midjourney, etc.
- LangChain framework for model I/O and orchestration, Python, Streamlit, etc.
- AWS Sagemaker Domain and Jumpstart API
The Impact
Delivering astounding business impact and embracing a future of endless creative opportunities
The generative AI program optimized campaign artifact generation for marketing and sales operations, paving the way for fast-tracked creativity and innovation.
- Content Generation Optimization – The novel content generation effort was optimized end-to-end. This empowered marketing teams to be more productive and freed up their time to focus on areas that needed their attention and expertise.
- First draft creation with Generative AI – 50-60% of all first drafts of brochures and other marketing operations were generated using generative AI. This reduced effort significantly slashed dependencies and overheads on several SaaS tools.
- Accelerated Go to Market – Launch and campaign timeframes were optimized by up to 30-40%, allowing for accelerated time to market.
- Innovation and Engagement - A Do It Yourself (D.I.Y) platform for the pharma commercial operations’ user community allowed for the simulation of creative pursuits, promoting the spirit of innovation and creativity across teams.
