| |
|
| |
| DETAILS |
|
Join us for an evening with Neeharika Tummala, Product Manager at Amazon, as she explores how product leaders are scaling LLM-powered systems. Neeharika will share how Alexa is evolving from a voice assistant into a more proactive, ambient AI experience.
As AI shifts from reactive tools to ambient systems, product management is becoming more complex. Neeharika Tummala will unpack how to choose models, manage trade-offs, & ensure reliability through observability & tuning. From architecture to user experience, this session explores how PMs can deliver both performance & delight at scale.
Neeharika Tummala is a Product Manager at Amazon, where she leads initiatives within the Alexa ecosystem focused on Voice AI, Ambient Computing, & Conversational AI. Her work sits at the intersection of complex machine learning infrastructure & user-centric design, specifically tasking LLMs with evolving from simple voice triggers into proactive, "delightful" web-based assistants.
LLMs are reshaping how products are built & experienced. Neeharika Tummala shares how product managers can move from voice-first systems to ambient AI, balance performance with reliability, & make decisions that drive both technical success & user delight.
01
The Strategy: Moving from Voice to Ambient Assistant
We will explore the logic of "expert orchestration," moving beyond simple voice commands to a cohesive web-based assistant. Neeharika will discuss some of the foundational challenges of working with Alexa's brain, the strategic framework for choosing a base model, evaluating performance trade-offs.
02
The Practice: Observability & Model Tuning
This session focuses on the technical application of LLMs within complex ecosystems. We will break down what observability looks like & how a model is refined for specific user intents, ensuring the output remains reliable as it scales across different platforms.
03
The Impact: Decision Making & User Delight
In an "Open the Veil" segment, we'll break down key development decisions. Attendees will see Alexa in action & critique PM choices, exploring how to balance constraints with a delightful user experience, plus the role of MCP, RAG, & OpenClaw in its strategy.
|
|
|
|
|
|
|
|
