The Hallway Track
ECRI named AI chatbot misuse the #1 health technology hazard of 2026, not a frontier problem but a production baseline issue.
“Most AI safety failures in health care are not model failures. They are architectural decisions that were made before even a single token was generated.”
DeepSeek 4.1 Flash outperforms previous models and significantly reduces operational costs.
“Incredible leap forward.”
AI's implementation isn't improving productivity in development teams.
“We used AI to solve a problem that AI created.”
Restate framework helps run resilient agents in production environments.
“We want to be able to bring it back and let it continue exactly at.”
Design skills are merging with engineering skills through tools like Impeccable.
“the role of the engineer and the design are blurring.”
AI has the potential to bridge the gap between design and engineering.
“I love AI.”
1Password improves engineering productivity by 21% using Codex.
Enterprise teams default to fine-tuning by accident, not conscious architectural choice
“most corporate teams make this decision by accident”
OpenAI publishes official GPT-6 family model guide for startup production workflows
Salesforce unveils vision for architecture in an agentic future.
AI verification and validation is crucial in chip development.
“it's like being in the 1990s, you have the internet, but they tell you that you can only use it from two to four o'clock.”
Databricks cut $1M annually in AI agent waste within a single hour of optimization work.
Models absorbing harness capabilities into weights is reshaping agent architecture toward human-attention scaffolding
“The change last winter, last Christmas — it's a little hard to pin down. I mean, the harness changed and a little post-training changed and then new pre-trained models came… but it felt like a big jump which is not that easy to pin down what did it.”
DeepMind's Gemma 4 achieves multimodal vision by patching images directly into the main transformer, eliminating separate encoders.
“Throw that all away. Out. Right now.”
Agent harness architecture drives double-digit benchmark swings independent of model choice
“Harness design alone can account for double-digit swings in benchmark results and significant differences in token cost”
Voice agents need sub-950ms response; small models beat frontier models on latency
“A frontier model that think for a full second has already lost the room, no matter how good the answer is.”
Sierra builds customer-engagement agents using many parallel models per turn and isolated PCI infrastructure for payments.
“We have isolated infrastructure where payment info doesn't go to an external large language model cuz none of the LLM providers are PCI certified in that way.”
Lyft scaled to seven+ production AI agents at 35% resolution by building an offline-eval quality gate before shipping.
“You don't want to use your users as test data.”
Traditional retry and circuit breaker patterns fail for LLMs; per-request provider fallback is required.
“If you have a single model provider, their ceiling is your ceiling. Their outage is your outage.”
AI architectural research fails because it tests at insufficient compute scale
“to get to any interesting results you need certain level of compute to even see the capabilities in the model”
LangChain's open-source Deep Agents adds production harness primitives to any tool-calling agent with one function swap
“if you're not the model, you're the harness”
Building production AI systems requires a four-phase framework—requirements, system design, evaluation, optimization—not just vibe coding.
“Specs are the new code. The art is in defining the product requirements, the system design, and evaluate criteria so you can be confident that your AI coding buddies are building the right thing.”
OpenGov built and scaled OG Assist, a production AI agent embedded across its government ERP product suite.
AWS combines NVIDIA NIM, Bedrock AgentCore, and Strands Agents for production-grade multi-agent systems
No good standards exist for agent development; real-world continual learning is largely absent.
“there aren't really good standards for these things, right? Like there there's not some one-size-fits-all uh solution”
GTM teams that bolt AI onto existing stacks cannot scale; architecture must be rebuilt with AI at the core.
“by the time the buyer reaches you, the decision is mostly made.”
Companies lose their fearless engineering culture and stop taking risks as they scale to thousands of employees.
“we would much rather fail in pursuit of the extraordinary than succeed in the ordinary”