Fully automated competitive monitoring and trend analysis
Integration of an LLM-powered chatbot for exploratory queries
Use of verified external sources to ensure high data quality
Technology stack
Agentic AI
Amazon Web Services
Azure OpenAI
Generative AI
LangChain
LLM Logging
LLM Orchestration
Palantir Foundry
Retrieval Augmentation Generation (RAG)
Vector database
„
The solution enabled us to fully automate the previously time-consuming competitive analysis. Thanks to Agentic AI, our client now gains strategically relevant insights much faster.
“
Christoph Bergen
CoE Lead for GenAI at HMS
The specialist departments of a global pharmaceutical company faced the challenge of preparing relevant market and competitive intelligence in a timely and structured way. To monitor news, research pipelines, acquisitions, and macroeconomic trends, employees had to manually collect, evaluate, and consolidate large volumes of data for management reports. These processes were time-consuming, error-prone, and tied up valuable resources, often taking several days to complete. Standard solutions could not meet the high requirements for data quality, timeliness, and scalability, creating a strong demand for an automated, flexible solution.
The HMS solution: Agentic AI for real-time competitive analysis
HMS developed a competitive intelligence platform that fully automates the entire competitor analysis process. An Agentic AI-based workflow autonomously orchestrates the selection of data sources, manages analysis processes, and triggers LLM calls. This provides specialist departments with up-to-date, structured information and enables them to perform additional analyses in real time – directly via a GenAI-powered chatbot.
Automated data collection and analysis
The solution aggregates reports from verified external, industry-relevant news sources. Initially, these reports are enriched with metadata such as category (e.g., research pipeline, acquisitions, macro trends) and sentiment. In subsequent steps, relevant developments, trends, and anomalies are identified. Under the hood, generative text models from Azure OpenAI are utilized, and the solution is designed to quickly integrate new models and techniques as they emerge.
The following diagram illustrates the key components and workflow of the solution:
Interactive reports and chatbot integration
Compact management reports are generated automatically. In addition, an LLM-powered chatbot allows users to ask ad-hoc questions and perform targeted analyses – without requiring technical expertise.
Automated data collection and analysis
The platform leverages the LangChain framework (LangChain, LangGraph, Langfuse) for orchestration and can easily be extended to integrate new data sources. It is cloud-based (AWS) and designed for global rollouts.
HMS reduced the analysis process from days to minutes.
