SPEAKER 2026

Aakriti Aggarwal
AI Research Engineer, IBM Research
About Talk
Switching Intelligence: Building Modular LLMs with Granite Switch
Attendees will be able to evaluate when a modular model architecture can be advantageous over a monolithic model and understand how adapters can be developed, combined, and utilized within the Granite Switch ecosystem.
TRACK: Developers' Meet
8th Oct 2026 | Hall B | Time:12:15-12:45
About Speaker
Aakriti Aggarwal is an AI Research Engineer at IBM Research, specialising in Applied AI, LLM inference, and enterprise Generative AI. She works on inference optimisation for IBM Granite foundation models and production deployments across hyperscaler infrastructure, with expertise in technologies such as vLLM and Red Hat Inference Server. An active open-source contributor, speaker, and AI community advisor, Aakriti has built Python tools for multi-agent systems, LLM serving, and agent orchestration. Her recognitions include Microsoft MVP in AI and being named among Favikon’s Top 200 AI Education and Productivity Creators globally in 2026.