Qualcomm

Accelerating AI on Edge — Chintan Parikh and Weiyi Wang, Google DeepMind

May 5, 2026 · 23:58

Chintan Parikh and Weiyi Wang from Google DeepMind present Gemma 4E edge models (2B and 4B) and the LiteRT framework for on-device AI, arguing that edge AI delivers latency, privacy, offline capability, and cost savings. The Gemma 4E models introduce agentic capabilities including built-in function calling, structured JSON output, and chain-of-thought reasoning, all optimized for hardware-native support across CPUs, GPUs, and NPUs. LiteRT supports cross-platform deployment on Android, iOS, macOS, Linux, Windows, and IoT devices like Raspberry Pi, with a CLI tool and AI Edge portal for benchmarking. Performance benchmarks show up to 56 tokens per second on iOS, 30x speedups with NPU acceleration, and 35x faster than Llama on mobile. The Gallery app demonstrates on-device skills such as Wikipedia querying, mood tracking, photo-to-music generation, and voice agents, with open-source code and a Hugging Face repository. Q&A addresses use cases like local security camera face recognition, LiteRT vs TensorRT on Orin, multi-agent architectures, and audio model support.

TLMs: Tiny LLMs and Agents on Edge Devices with LiteRT-LM — Cormac Brick, Google

May 3, 2026 · 1:20:58

Cormac Brick from Google AI Edge explains how Tiny LLMs (sub-1B parameters) and on-device agent skills are making edge AI practical. He details LiteRT-LM, an open-source runtime that runs Gemma models on Android, iOS, and embedded systems, achieving over 1,000 tokens/s on high-end phones. Agent skills use progressive disclosure—loading skill details on demand—enabling reliable tool calling on 2B-4B models. For app deployment, fine-tuning boosts tiny model reliability by 20-40 points (e.g., Function Gemma 270M hit 86% accuracy on 10 functions). Synthetic data workflows and modular design (ASR + text polishing) power real apps like AI Edge Eloquent, which runs entirely offline. Safety is managed through system-level checkers and narrow functional scope for tiny models.

Forget RAG Pipelines—Build Production Ready Agents in 15 Mins: Nina Lopatina, Rajiv Shah, Contextual

Jun 27, 2025 · 1:15:43

Rajiv Shah, Nina Lopatina, and Matthew from Contextual AI demonstrate how developers can build production-ready RAG agents in minutes using Contextual AI's managed RAG platform, emphasizing that RAG should be treated as a managed service to avoid reinventing infrastructure. They walk through ingesting documents like NVIDIA financials and spurious correlation reports, then querying the agent with questions requiring quantitative reasoning across tables. The platform handles extraction, layout analysis, image captioning, hybrid retrieval, and a state-of-the-art reranker, capped by a grounded language model that avoids hallucinations and provides attribution. Evaluation is done via LM Unit, a model-as-judge that scores responses on criteria like accuracy and causation. The episode also shows integrating the agent with Claude Desktop via MCP and answers audience questions on pricing (consumption-based with a $25 credit), scalability, entitlements, HIPAA, and domain-specific language.

Foundry Local: Cutting-Edge AI experiences on device with ONNX Runtime/Olive — Emma Ning, Microsoft

Jun 27, 2025 · 22:52

Emma Ning, Principal PM at Microsoft, presents Foundry Local, a platform for building cross-platform on-device AI applications using ONNX Runtime and Olive. She argues local AI is essential for low-bandwidth, privacy-sensitive, cost-efficient, and real-time latency scenarios, noting that modern hardware and optimized models like Phi-4-mini and DeepSeek make it viable. Ning demonstrates Foundry Local's CLI, model benchmarking (Qwen 1.5B at 90 tok/s vs. Phi-4-mini with richer output), and a cross-platform document summarization app built with JavaScript SDK, running identically on Windows and macOS. She also previews a local agent with MCP servers (file system and OCR) that extracts receipt totals using Phi-4-mini. Customer testimonial from Sava (CEO of Pieces) highlights improved memory management and tokens-per-second, while another partner underscores ease of install and hybrid cloud-local solutions.

RAG Agents in Prod: 10 Lessons We Learned — Douwe Kiela, creator of RAG

Apr 10, 2025 · 16:56

Douwe Kiela, CEO of Contextual AI and creator of RAG, shares 10 lessons from deploying enterprise RAG systems at scale. He argues that language models are only 20% of a larger system; success comes from focusing on systems, not models, and specializing over AGI to unlock domain expertise. Enterprise data is the real moat, but pilots are easy while production is hard—design for production from day one. Speed beats perfection: ship barely functional to real users early and iterate. Avoid boring engineering chores like chunking; instead, integrate AI into existing workflows to drive adoption. Accuracy is table stakes; handle inaccuracy with observability and attribution. Be ambitious: aim for transformative ROI, not low-hanging fruit like basic HR questions.