What's New in Inference Engineering
Conference Context
- Date/time: 2026-07-01 · 1:30pm-1:50pm
- Track/room: Inference · Track 9
- Speaker(s): Philip Kiely
- Session type/status: session · confirmed
- Track: Inference
- Room: Track 9
- Session type: session
- Status: confirmed
Session Description
More than 30,000 engineers have learned the fundamentals of inference since Inference Engineering was published. But the field keeps accelerating, so it's time for the first public addendum to the book. The past four months have seen a renewed focus on training-dependent inference optimization across the "big three" performance techniques of speculation, caching, and quantization. This talk provides structured guidance for training DFlash and EAGLE 3 draft models to accelerate LLM decode, introduces the concept of KV compaction, and explains the hype behind TurboQuant.
Media Evidence
Optimizing inference for voice models in production - Philip Kiely, Baseten (speaker-match related prior/adjacent AI Engineer video; captions: English auto-captions).
- Source video:
youtube-gmTHs5T_YAE - Slide deck: Dense Slides: Optimizing inference for voice models in production - Philip Kiely, Baseten — 2 visible slide image(s); 2 HTML recreation(s).
- Additional slide evidence: Slides: Optimizing inference for voice models in production - Philip Kiely, Baseten, Reconstructed Slides: Optimizing inference for voice models in production - Philip Kiely, Baseten

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-gmTHs5T_YAE— source page linked- Evidence links for
youtube-gmTHs5T_YAE: youtube gmTHs5T_YAE, youtube gmTHs5T_YAE slides, youtube gmTHs5T_YAE dense slides, youtube gmTHs5T_YAE 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
Supporting Slides
- youtube gmTHs5T_YAE slides — extracted from the related public AI Engineer video.
Slide Evidence
- Slide-only cropped deck: youtube gmTHs5T_YAE dense slides (3 viable slide images).
- Related slide/OCR pages:
- youtube gmTHs5T_YAE dense slides
- youtube gmTHs5T_YAE reconstructed slides
- youtube gmTHs5T_YAE slides
- Slide-derived terms:
orpheus,microsoft,baseten,model,architecture,performance,graphite,windsurf,mongobb,mdaily,augmentcode,workos,philip,developer,relations,based,sfba,favorite
Synthesis
Synthesized Breakdown
What's New in Inference Engineering ## Conference Context - Date/time: 2026-07-01 · 1:30pm-1:50pm - Track/room: Inference · Track 9 - Speaker(s): Philip Kiely - Session type/status: session · confirmed - Track: Inference - Room: Track 9 - Session type: session - Status: confirmed ## Session Description More than 30,000 engineers have learned the fundamentals of inference since Inference Engineering was published. But the field keeps accelerating, so it's time for the first public addendum to the book. The past four months have seen a renewed focus on training-dependent inference optimization across the "big three" performance techniques of speculation, caching, and quantization. This talk provides structured guidance for training DFlash and EAGLE 3 draft models to accelerate LLM decode, introduces the concept of KV compaction, and explains the hype behind TurboQuant.
Speaker And Company Context
- Philip Kiely — Developer Relations at Baseten.
Topics Covered
Derived Links And Source Material
- youtube gmTHs5T_YAE — related YouTube source page.
- youtube gmTHs5T_YAE slides — slide evidence.
- youtube gmTHs5T_YAE reconstructed slides — slide evidence.
- youtube gmTHs5T_YAE 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.