Physical AI at GSK: Unlocking business value with foundation models and digital twins
When GSK set out to serve 1.4 billion patients annually across 37 manufacturing sites worldwide, with a goal of reaching 2.5 billion by 2031, Sander Timmer and his AI team faced a daunting reality: the pharmaceutical supply chain is one of the most complex decision networks on earth. Making medicines and vaccines takes years, cycles are long, and every decision has massive downstream impacts that ripple across continents.
“It’s not easy,” said Timmer. “You can’t make decisions in silos. You need to integrate the entire complex decision network, understand KPIs, map them together, and feed them to agentic systems so you can simulate the impact of decisions.”
What started as an effort to move beyond static BI dashboards has evolved into something far more ambitious: a comprehensive “Physical AI” strategy that spans agentic supply chains, real-time digital twins controlling living biological processes, and an AI infrastructure foundation supporting cutting-edge innovation across global manufacturing sites and complex workflows.
At the center of this transformation? Weights & Biases, serving as the unified platform that brings order to GSK’s sprawling multi-cloud, multi-vendor AI ecosystem spanning thousands of production models.
Read on to learn how GSK has built the world’s most sophisticated pharmaceutical AI operations platform, and why W&B’s Weave and Models became indispensable to making it all work.
The impossible complexity of pharmaceutical AI
GSK’s AI challenges are unique in their scale and consequences. The company processes demand planning and forecasting, sourcing and procurement, logistics and transportation management, warehouse management, inventory optimization, and final mile delivery—all while maintaining the stringent quality controls and regulatory compliance that pharmaceutical manufacturing demands.
The team needed to overcome several critical challenges that made traditional AI approaches insufficient:
Zero tolerance for error: In pharmaceuticals, false negatives are absolutely unacceptable. When AI makes a decision affecting patient safety or drug quality, the basis and history of that decision must be recorded more precisely than human decision-making.
Extreme operational complexity: With manufacturing sites using Google Cloud, Microsoft Cloud, Databricks, NVIDIA Edge, and multiple other vendors, getting a unified view of AI operations across different platforms, sources and deployment environments was extremely challenging.
The “black box” problem in regulated industries: Without comprehensive tracing, it’s impossible to understand why models produced specific outcomes. For audits, inspections, and continuous improvement, GSK needed complete visibility into every AI decision.
Constant foundation model evolution: The rapid pace of new LLMs and frameworks creates continuous regression risk—new features can unintentionally break existing agent capabilities and degrade performance in production.
Moving from reactive to proactive MLOps: GSK’s initial approach was entirely manual and reactive. Nine sites around the world ran local inference, stored results in a central database that humans validated, and reviewed accuracy in monthly PDF reports. When problems emerged, it could take five months from discovery to fix deployment.
Building the Agentic Supply Chain
Rather than continue relying on BI dashboards where humans manually investigate every anomaly, Sander and his team envisioned a fundamentally different paradigm: agentic systems that work with data, not just report on it.
“The previous approach was: person in charge views KPIs on a dashboard, if there’s an abnormality someone checks by phone or email, then makes decisions by mobilizing experience and intuition,” explained Timmer. “Now, agents automatically detect anomalies, present root cause hypotheses to the brand manager, simulate scenarios with downstream impacts, and people can focus on choosing which action to adopt.”
To power this vision, GSK collaborated with Microsoft to build AIGA, an internal framework for GenAI and agentic solutions. Rather than simply connecting LLMs, AIGA orchestrates a sophisticated flow: searching short-term memory, meeting notes, and standard operating procedures; planning which apps, tools, and data to access; then moving to inference and answer generation—with every step fully traceable.
But building agents that work in development is vastly different from running them reliably in production at pharmaceutical scale.
LQAI: The world’s first fully AI-managed ETF and the EXAONE-GanphA agentic framework supporting it
The LQAI ETF represents a breakthrough in AI-driven investment management. Rebalancing every four weeks, the fund uses AI to select 100 stocks from the S&P 500, with decisions made entirely through their EXAONE-GanphA framework.
The system processes over 5,000 news articles daily, training continuously on market information to generate investment signals. The results speak for themselves: LQAI delivered 27.82% performance compared to SPY’s 24.89% and Goldman Sachs Active Beta Large Cap ETF’s 24.21%. The ETF also has a monthly generated explanatory report for the decisions it makes.
“Think of it as Warren Buffet’s annual letter, but generated entirely by AI,” said Choi. “Having the LLM generate explanations and reasoning for forecast results in natural language enhances understanding and trust in the AI decision making.”
The technical architecture behind these impressive results and outputs is EXAONE-GanphA, a modular AI framework designed to model, explain, and act within dynamic environments. The system employs four complementary AI agents working in concert:
AI Journalist: Curates timely market narratives by processing thousands of news sources daily, identifying relevant themes and sentiment patterns that traditional quantitative models miss.
AI Economist: Forecasts markets with full context, producing forward-looking signals by combining time-series modeling with exogenous conditioning and dynamic variable interaction analysis.
AI Analyst: A generative-explanatory module that translates numerical outputs into structured narratives, combining multi-source data fusion with scenario modeling and LLM-based explanation generation.
AI Decision-Maker: Turns insights into executable decisions, operationalizing insights through algorithmic ranking, optimization, and feedback simulation—whether for ETF rebalancing, supply chain optimization, or strategic planning.
This architecture is supported by three key technical innovations that LG AI Research developed through their forecasting projects:
Multi-modal deep document understanding (DDU): Transforms unstructured documents into formats that AI can comprehend and utilize, enabling the system to process corporate filings, research reports, and regulatory documents alongside traditional market data.
Advanced time-series analysis: Uses deep learning models like Transformers with mixed frequency, multi-variate, and ensemble learning methods to capture complex temporal patterns.
LLM-powered explainability: The same language model that processes market information generates explanations and reasoning for forecast results in natural language, solving the black box problem that has limited AI adoption in financial services.
Throughout the development of both EXAONE and the forecasting applications, Weights & Biases has been integral to LG AI Research’s success.
“We leverage heavily the product suite of Weights & Biases to build our models and fine-tune them,” said Choi. “There are many fine-tuning needs across all our affiliate companies, and we’re using it very heavily right now and are very happy with the product.”
The future of AI in financial services
LG AI Research’s work represents a fundamental shift in how AI can be applied to financial forecasting. By combining the power of large language models with traditional quantitative methods, they’ve created a system that doesn’t just predict—it explains, adapting to new information while providing the transparency that financial institutions require.
“We’re really excited to be pioneers in this space and lead the way,” said Choi. “Now that we have context from news and text, you’re also able to explain the decision making process and what the rationale is behind the outputs that AI’s are generating, which can potentially solve the Black Box issue which exists in financial services.”
As LG AI Research continues expanding their capabilities—from molecular structure prediction with EXAONE Discovery to pathological image analysis with EXAONE Path—their financial forecasting framework represents just the beginning of what’s possible when cutting-edge AI research meets real-world business challenges.
Learn more about the LQAI ETF here.
