---
title: "Model Routing"
category: "talks"
date: "2026-07-01"
time: "3:45pm-4:05pm"
track: "Local AI"
room: "Track 4"
speakers: ["Nader Khalil", "Walden Yan", "Tanay Varshney", "Alex Atallah"]
sourceLabels: ["Official conference schedule", "Public YouTube metadata"]
scheduleTrack: "Local AI"
scheduleRoom: "Track 4"
scheduleLabels: ["Local AI", "Track 4", "session", "confirmed"]
---
# Model Routing

## Conference Context
- Date/time: 2026-07-01 · 3:45pm-4:05pm
- Track/room: Local AI · Track 4
- Speaker(s): Nader Khalil, Walden Yan, Tanay Varshney, Alex Atallah
- Session type/status: session · confirmed

- Track: Local AI
- Room: Track 4
- Session type: session
- Status: confirmed

## Session Description
Model Routing explores how teams decide when to use local models, open-source models, or frontier cloud systems, and why the answer is increasingly hybrid rather than one-size-fits-all. The panel digs into routing architectures, model selection strategies, stack decisions, and what still needs to improve in local AI before more workloads can move closer to the user. Moderator: Nader Khalil (NVIDIA). Panelists: Walden Yan (Cognition), Tanay Varshney (NVIDIA), Alex Atallah (OpenRouter).

## Media Evidence
[Fun stories from building OpenRouter and where all this is going - Alex Atallah, OpenRouter](https://www.youtube.com/watch?v=84Vtz2IL1Ug) (speaker-match related prior/adjacent AI Engineer video; captions: English auto-captions).

- Source video: `youtube-84Vtz2IL1Ug`
- Slide deck: [[youtube-84Vtz2IL1Ug-dense-slides|Dense Slides: Fun stories from building OpenRouter and where all this is going - Alex Atallah, OpenRouter]] — 2 visible slide image(s); 2 HTML recreation(s).
![[assets/dense-slides/84Vtz2IL1Ug/slide-001.jpg]]
![[assets/dense-slides/84Vtz2IL1Ug/slide-002.jpg]]
- Additional slide evidence: [[youtube-84Vtz2IL1Ug-slides|Slides: Fun stories from building OpenRouter and where all this is going - Alex Atallah, OpenRouter]], [[youtube-84Vtz2IL1Ug-reconstructed-slides|Reconstructed Slides: Fun stories from building OpenRouter and where all this is going - Alex Atallah, OpenRouter]]
- Slide-derived themes for `youtube-84Vtz2IL1Ug`: llama, open, efficient, foundation, language, models, alpaca, strong.

## Evidence Graph
This evidence graph is generated from currently linked source material: official schedule text, related video pages, cached transcripts, visible slide text, dense/reconstructed slide pages, and AI slide-classification audits.

### Media Signals
- `youtube-84Vtz2IL1Ug` — 2 slide-derived text signals
- Slide-derived themes for `youtube-84Vtz2IL1Ug`: llama, open, efficient, foundation, language, models, alpaca, strong.
- Evidence links for `youtube-84Vtz2IL1Ug`: [[youtube-84Vtz2IL1Ug]], [[youtube-84Vtz2IL1Ug-slides]], [[youtube-84Vtz2IL1Ug-dense-slides]], [[youtube-84Vtz2IL1Ug-reconstructed-slides]]

### Agent Reading Notes
Use these signals to refine the synopsis, topic links, people/company context, and method notes. If a source is a related external video rather than an exact official recording, keep it framed as supporting evidence.

## Transcript Status
Related video transcript availability: English auto-captions. Treat this as supporting context, not a recording of this exact scheduled session unless later confirmed. Not fetched yet.

## People
- [[nader-khalil]]
- [[walden-yan]]
- [[tanay-varshney]]
- [[alex-atallah]]

## Slide Evidence
- Slide-only cropped deck: [[youtube-84Vtz2IL1Ug-dense-slides]] (3 viable slide images).
- Related slide/OCR pages:
- [[youtube-84Vtz2IL1Ug-dense-slides]]
- [[youtube-84Vtz2IL1Ug-reconstructed-slides]]
- [[youtube-84Vtz2IL1Ug-slides]]
- Slide-derived terms: `models`, `tokens`, `world`, `fair`, `model`, `llama`, `language`, `inference`, `train`, `openrouter`, `engineering`, `future`, `performance`, `openal`, `best`, `training`, `datasets`, `alengincer`

## Synthesis
### Synthesized Breakdown
# Model Routing ## Conference Context - Date/time: 2026-07-01 · 3:45pm-4:05pm - Track/room: Local AI · Track 4 - Speaker(s): Nader Khalil, Walden Yan, Tanay Varshney, Alex Atallah - Session type/status: session · confirmed - Track: Local AI - Room: Track 4 - Session type: session - Status: confirmed ## Session Description Model Routing explores how teams decide when to use local models, open-source models, or frontier cloud systems, and why the answer is increasingly hybrid rather than one-size-fits-all. The panel digs into routing architectures, model selection strategies, stack decisions, and what still needs to improve in local AI before more workloads can move closer to the user. Moderator: Nader Khalil (NVIDIA). Panelists: Walden Yan (Cognition), Tanay Varshney (NVIDIA), Alex Atallah (OpenRouter).

### Speaker And Company Context
- [[nader-khalil|Nader Khalil]] — Director of Developer Technology at [[nvidia|NVIDIA]].
- [[walden-yan|Walden Yan]] — Co-founder & CPO at [[cognition|Cognition]].
- [[tanay-varshney|Tanay Varshney]] — Principal Engineer at [[nvidia|NVIDIA]].
- [[alex-atallah|Alex Atallah]] — Co-founder & CEO at [[openrouter|OpenRouter]].

### Topics Covered
- [[coding-agents]]

### Derived Links And Source Material
- [[youtube-84Vtz2IL1Ug]] — related YouTube source page.
- [[youtube-84Vtz2IL1Ug-slides]] — slide evidence.
- [[youtube-84Vtz2IL1Ug-reconstructed-slides]] — slide evidence.
- [[youtube-84Vtz2IL1Ug-dense-slides]] — slide evidence.

### Novel Concepts / Clever Methods
- No highlighted novel concept has been detected yet.

### Evidence Boundary
This synthesis is based on the official schedule and linked source pages. It should be revisited when exact session recordings or transcript-backed secondary sources are available.