Fireside Chat: Innovation Efficiency

Singapore AI & Robotics Demo Nights aim to bring together builders, founders and AI researchers together to share insights, challenges and breakthroughs. This is a night for inspiring demonstrations of AI capabilities. Come for this event if you want to catch a peek into the future of this fast growing industry. Event Details: 🕒 7pm-8.30pm (doors open 6.30pm) 📍 National Library Board, Level 7 - Launch Programme Room Demo Showcase Raghav, Integral, Downloading more RAM for my life - Hermes/OpenClaw beyond the hype Ankit & Febby, Amorphous Robotics Pte Ltd, Multimodal Haptic Feedback for Virtual Interactions Hanna, Ohannas Studio, From Drawing to Reality Lee Gang, ELGO AI, Evaluating Multi-Agent Systems Rain, Autodemia, Autodemia AI Research Automation Platform if you are interested in giving a demo, please fill in this form Agenda 6:30 PM: Doors open & networking 7:00 PM: Lightning Talks begin! Note: It's a library, no food provided so please eat before/after the event! Who Should Attend AI Practitioners and Engineers interested in technical aspects of AI model training & optimization Businesses and Decision-makers looking to understand the latest capabilities in AI Investors, Industry Analysts interested in tracking trends in AI development and infrastructure

Returning for its 11th edition, TechLaw.Fest 2026 marks a milestone year, coinciding with the bicentenary of Singapore’s modern legal system. At this legaltech conference, we will examine how core principles of fairness, accountability and access to justice can be upheld amid rapid technological change. Bringing together over 2,000 participants from more than 30 countries, TechLaw.Fest convenes legal professionals, technologists, policymakers and academics to exchange ideas, spark collaboration and shape the future of technology law and digital transformation. Join us in discovering how the practice of law and delivery of legal services will evolve and transform between East and West! 🗓️ 9-10 September 2026 📍 Sands Expo and Convention Centre, Singapore Get your complimentary Trade Visitor Pass. To enjoy 40% savings on the 2-day conference pass for full access to all stages, use promo code INTL700 during checkout. Register now at https://tinyurl.com/TLF26xACESG

We’re back on 9 September for the second Singapore Hardware Meetup, bringing together the people building physical products across Singapore. Founders, engineers, designers, investors, and hardware enthusiasts are all welcome. Hardware Meetups take place in more than 40 cities worldwide, connecting hardware professionals to meet, collaborate, and learn from one another. This Edition: From Prototype to Production Building a working prototype is often only the beginning. Turning it into a reliable product—and scaling production—is where hardware startups succeed or fail. Three speakers from across Singapore’s hardware ecosystem will share practical lessons from their journeys, based on what they are building today. And because this is a Hardware Meetup, don’t come empty-handed. Bring your prototypes, demos, and inventions to showcase during the networking session. Speakers Eric Cao Co-Founder and CEO, Qifrost Qifrost is an A*STAR Quantum Innovation Centre spin-out building cryogenic control systems for quantum computing. Yip Jia Qi Senior Research Engineer, Menlo Research Menlo Research is a Singapore physical AI company developing open-reference hardware for humanoid robotics, including the open-source Asimov humanoid. Adriel Ng Co-Founder and COO, EcoFlow EcoFlow is a Singapore water technology company deploying smart water monitoring and conservation solutions across hotels and commercial buildings in Southeast Asia. Programme 3:30–4:00 PM — Registration 4:00–4:10 PM — Welcome and introduction 4:10–4:25 PM — Eric Cao, Qifrost 4:25–4:40 PM — Yip Jia Qi, Menlo Research 4:40–4:55 PM — Adriel Ng, EcoFlow 4:55–5:00 PM — Closing 5:00–6:30 PM — Networking and product demonstrations About the Host Factorem is an end-to-end manufacturing partner for custom parts. Through AI-powered instant quoting, design-for-manufacturing reviews, and a vetted factory network covering CNC machining, sheet metal, 3D printing, and injection moulding, Factorem helps hardware teams move from p

About the event Building production AI products means two hard problems at once: a data layer fast enough to power real-time analytics, and LLM calls observable enough to debug at scale. Most teams stitch together a separate analytics database, a warehouse, an observability stack, a vector store, and a tracing tool — and the seams are where production AI breaks. Join ClickHouse and the Langfuse team for an evening on the Agentic Data Stack — one columnar engine for analytics, observability, Postgres, and agents. The payoff: a unified stack and lower total cost of running AI in production. What we'll cover The Agentic Data Stack — ClickHouse’s vision for a unified foundation for production AI: real-time analytics, observability (ClickStack), Postgres (ClickHouse Postgres, in public beta), and agents (managed agentic analytics powered by Claude) — all on one columnar engine. One stack, one billing line, one telemetry surface; less data duplication, fewer pipelines to operate. Langfuse — open-source LLM observability, now part of the ClickHouse family. Trace every LLM call, tool invocation, retrieval step, and agent action with cost, latency, and quality analytics — on a ClickHouse-native data model that scales to hundreds of millions of observations. You'll walk away with A clear mental model for the Agentic Data Stack: how ClickHouse, ClickStack, ClickHouse Postgres, and ClickHouse Agents combine into one engine — replacing the fragmented analytics / warehouse / observability / vector patchwork with a single, unifying data layer. How to instrument an LLM application with Langfuse in under 30 minutes — open-source, OpenTelemetry-based, with 100+ framework integrations (LangChain, LlamaIndex, LiteLLM, OpenAI, and more) and Python and JS/TS SDKs. A reference architecture for production AI on the Agentic Data Stack: real-time analytics + LLM observability on the same columnar engine — unifying the stack while cutting total cost. Practical patterns for monitoring cost, la

This workshop goes from raw recording to curated corpus. We'll cover what MCAP is and why it's built that way, tour real Physical AI datasets across driving, aquatic, and forest robots, and open an episode with every sensor synced—including channels nothing knows how to decode. Time, Date and Location Sep 09, 2026 9:00 AM - 10:00 AM PST Online. Register for the Zoom! Physical AI still relies on familiar computer vision tasks—detection, segmentation, depth, tracking. What's changed is the data unit: no longer a single image and label, but an episode—a dozen sensors ticking on independent clocks for minutes, with no frame boundaries. Most computer vision tooling assumes the old unit and breaks on the new one. That's why Physical AI teams end up with buckets of .mcap files nobody can characterize. Recording is cheap, so logs pile up faster than anyone curates them. Ask what's actually in there—which tasks, which conditions, how many failures and of what kind—and the honest answer is usually a shrug. MCAP has been ROS 2's default log format since Iron, and as of FiftyOne 1.19 it opens natively: cameras, LiDAR, GPS, IMU, and logs on one shared timeline, alongside your images and video. We'll tackle quality, the harder half: what smoothness, sensor-health, and outlier metrics actually measure, where each falls short, and how to turn a score into a defensible decision. You'll leave knowing how to load your own recordings, query a whole corpus instead of a single file, and which quality signals to trust for which job.