Crusoe hosts inference roundtable May 28

Published May 27, 2026 by The Daily Scout

- Crusoe said it will host a May 28 roundtable in London focused on training and inference optimization, with SVP of Product Erwan Menard speaking. - The event listing says arrivals begin at 08:30 and the discussion will cover managed inference, Crusoe Cloud, and production best practices with Earlybird's Akash Bajwa. - May 28’s session is listed on Crusoe’s events page and Luma, with Menard and Bajwa named participants.

Why it matters

Crusoe is turning a product pitch into a live operator discussion this week. The AI infrastructure company has scheduled a May 28 roundtable in London centered on training and inference optimization, according to Crusoe’s events page and a Luma registration listing. The session features Erwan Menard, Crusoe’s senior vice president of product management, alongside Akash Bajwa, a principal at Earlybird. The agenda points to managed inference, Crusoe Cloud and the practical work of running models in production. ### Who is speaking, and when does it start? May 28 is the date listed on Crusoe’s events page for “Crusoe AI Talk: London,” with timing shown as 5 p.m. to 9 p.m. BST. A separate Luma page for the event says arrivals begin at 08:30 and run until 09:00, followed by a discussion on training and inference optimization. Erwan Menard is named on the Luma listing as the Crusoe speaker. Akash Bajwa is identified on Earlybird’s website as a principal based in London who focuses on software, applied AI, infrastructure, models and deep tech. ### What does Crusoe say the discussion will cover? The Luma listing says the session will focus on “best practices for training and inference optimisation.” It describes Crusoe as offering managed inference through Crusoe Cloud and says attendees should expect discussion with founders and operators building in AI. (crusoe.ai) Crusoe’s own product materials show why those topics are central to its current sales message. (luma.com) The company says Crusoe Managed Inference is built to provide low latency, resilient scaling and predictable cost for open and open-source model inference on Crusoe Cloud. ### Why is managed inference at the center of Crusoe’s pitch? November 20, 2025 is when Crusoe announced the general availability of Crusoe Managed Inference. (luma.com) In that release, the company said the service was designed for production AI workloads and claimed gains of up to 9.9 times faster time-to-first-token and up to 5 times higher throughput. (crusoe.ai) April 8, 2026 is when Crusoe published a blog post saying inference had overtaken training as the dominant AI workload. Menard said in that post that inference is now “the biggest workload — bigger than training — across all clouds consolidated,” framing inference cost and reliability as the main infrastructure problem for teams deploying AI products at scale. (crusoe.ai) ### How does this connect to Crusoe Cloud? March 2025 was when Crusoe announced new AI platform services on Crusoe Cloud, including Managed Inference and AutoClusters, as part of its push to offer managed deployment services on top of NVIDIA-accelerated infrastructure. February 2026 was when Menard wrote that Crusoe customers could bring their own models to Managed Inference, including models trained from scratch or tuned from open models. (crusoe.ai) In that post, he said customers were benchmarking price and performance against existing baselines and that Crusoe had seen multiple cases of more than 4 times gains in queries per second with lower latency. (crusoe.ai) ### Why include a venture investor in an infrastructure roundtable? Akash Bajwa’s profile at Earlybird says he works with early-stage companies across software, fintech, applied AI, infrastructure and models. That makes him a relevant outside participant for a discussion aimed at founders and operators weighing cloud choices, deployment trade-offs and infrastructure costs. The Luma listing says the event is intended as an “honest discussion” with a curated group of builders. (crusoe.ai) Crusoe has used similar event formats in recent weeks, including other talks and meetups around AI infrastructure and open-source model deployment. May 28 is the next dated milestone. Crusoe’s events page lists the London session for that day, and the Luma registration page names Menard and Bajwa as participants in the discussion on training and inference optimization. (earlybird.com) (crusoe.ai) (luma.com)

Key numbers

Crusoe said it will host a May 28 roundtable in London focused on training and inference optimization, with SVP of Product Erwan Menard speaking.
The event listing says arrivals begin at 08:30 and the discussion will cover managed inference, Crusoe Cloud, and production best practices with Earlybird's Akash Bajwa.
May 28’s session is listed on Crusoe’s events page and Luma, with Menard and Bajwa named participants.
The AI infrastructure company has scheduled a May 28 roundtable in London centered on training and inference optimization, according to Crusoe’s events page and a Luma registration listing.

What happens next

The AI infrastructure company has scheduled a May 28 roundtable in London centered on training and inference optimization, according to Crusoe’s events page and a Luma registration listing.
May 28 is the date listed on Crusoe’s events page for “Crusoe AI Talk: London,” with timing shown as 5 p.m.
A separate Luma page for the event says arrivals begin at 08:30 and run until 09:00, followed by a discussion on training and inference optimization.

Sources

Quick answers

What happened in Crusoe hosts inference roundtable May 28?

Crusoe said it will host a May 28 roundtable in London focused on training and inference optimization, with SVP of Product Erwan Menard speaking. The event listing says arrivals begin at 08:30 and the discussion will cover managed inference, Crusoe Cloud, and production best practices with Earlybird's Akash Bajwa. May 28’s session is listed on Crusoe’s events page and Luma, with Menard and Bajwa named participants.

Why does Crusoe hosts inference roundtable May 28 matter?

Crusoe is turning a product pitch into a live operator discussion this week. The AI infrastructure company has scheduled a May 28 roundtable in London centered on training and inference optimization, according to Crusoe’s events page and a Luma registration listing. The session features Erwan Menard, Crusoe’s senior vice president of product management, alongside Akash Bajwa, a principal at Earlybird. The agenda points to managed inference, Crusoe Cloud and the practical work of running models in production. Who is speaking, and when does it start? May 28 is the date listed on Crusoe’s events page for “Crusoe AI Talk: London,” with timing shown as 5 p.m. to 9 p.m. BST. A separate Luma page for the event says arrivals begin at 08:30 and run until 09:00, followed by a discussion on training and inference optimization. Erwan Menard is named on the Luma listing as the Crusoe speaker. Akash Bajwa is identified on Earlybird’s website as a principal based in London who focuses on software, applied AI, infrastructure, models and deep tech. What does Crusoe say the discussion will cover? The Luma listing says the session will focus on “best practices for training and inference optimisation.” It describes Crusoe as offering managed inference through Crusoe Cloud and says attendees should expect discussion with founders and operators building in AI. (crusoe.ai) Crusoe’s own product materials show why those topics are central to its current sales message. (luma.com) The company says Crusoe Managed Inference is built to provide low latency, resilient scaling and predictable cost for open and open-source model inference on Crusoe Cloud. Why is managed inference at the center of Crusoe’s pitch? November 20, 2025 is when Crusoe announced the general availability of Crusoe Managed Inference. (luma.com) In that release, the company said the service was designed for production AI workloads and claimed gains of up to 9.9 times faster time-to-first-token and up to 5 times higher throughput. (crusoe.ai) April 8, 2026 is when Crusoe published a blog post saying inference had overtaken training as the dominant AI workload. Menard said in that post that inference is now “the biggest workload — bigger than training — across all clouds consolidated,” framing inference cost and reliability as the main infrastructure problem for teams deploying AI products at scale. (crusoe.ai) How does this connect to Crusoe Cloud? March 2025 was when Crusoe announced new AI platform services on Crusoe Cloud, including Managed Inference and AutoClusters, as part of its push to offer managed deployment services on top of NVIDIA-accelerated infrastructure. February 2026 was when Menard wrote that Crusoe customers could bring their own models to Managed Inference, including models trained from scratch or tuned from open models. (crusoe.ai) In that post, he said customers were benchmarking price and performance against existing baselines and that Crusoe had seen multiple cases of more than 4 times gains in queries per second with lower latency. (crusoe.ai) Why include a venture investor in an infrastructure roundtable? Akash Bajwa’s profile at Earlybird says he works with early-stage companies across software, fintech, applied AI, infrastructure and models. That makes him a relevant outside participant for a discussion aimed at founders and operators weighing cloud choices, deployment trade-offs and infrastructure costs. The Luma listing says the event is intended as an “honest discussion” with a curated group of builders. (crusoe.ai) Crusoe has used similar event formats in recent weeks, including other talks and meetups around AI infrastructure and open-source model deployment. May 28 is the next dated milestone. Crusoe’s events page lists the London session for that day, and the Luma registration page names Menard and Bajwa as participants in the discussion on training and inference optimization. (earlybird.com) (crusoe.ai) (luma.com)