Reliable AI Forum #5: Not All Answers Are Equal
An evening for data science leaders evaluating the next generation of AI tools
Natural language interfaces are everywhere—from dashboards to embedded copilots. But when the questions are serious, the stakes are high, and the data is messy, accuracy isn’t just a feature—it’s a requirement.
In the 5th edition of Reliable AI Forum, we explore what it really takes to trust AI systems that talk to your internal data. These “data agents” promise insight and automation, but too often fail under real-world complexity: ambiguous queries, multi-system logic, shifting schema, and business-critical metrics.
Join us for a high-signal evening of insight, research, and demos—all designed for data science leaders tasked with evaluating or building the next generation of trustworthy AI.
🧠 What to Expect
Expect deep dives into:
Frameworks for evaluating accuracy in natural language AI tools used on internal, high-stakes data
Lessons from applied science leadership at Uber and Lightspeed on what makes AI trustworthy in practice
New research from PromptQL Labs on building deterministic, agentic reasoning systems
Some exciting demos!
🗓️ Agenda
5:00 PM – Doors open
5:30–6:15 PM – Fireside Chat with Eoin O’Mahony (Partner, Data Science @ Lightspeed, ex-Uber)
Not All Answers Are Equal: How Data Science Leaders Evaluate Natural Language AI
A candid conversation on how data science leaders can evaluate natural language AI systems—especially when reliability and precision matter. Eoin will share insights from scaling applied science at Uber, and how he’s now helping shape the future of AI infrastructure at Lightspeed to support high-stakes, data-driven decision-making.
6:15–7:00 PM – PromptQL Labs Research Showcase
What It Takes to Build Trustworthy Natural Language Systems
Explore new methods from the PromptQL research team tackling reasoning reliability, multi-system queries, and evaluation frameworks that reflect the real-world complexity data teams face.
7:00–7:20 PM – Live Demos
Short, high-signal demos from builders! Submit here!
Confirmed demos to be announced soon.
7:30–8:30 PM – Dinner, drinks & networking
8:30 PM – Doors close
→ If you'd like to speak at this meetup, you can submit your talk here.
About Reliable AI Forum
A monthly series for data and AI leaders exploring how to build systems that deliver accuracy, repeatability, and trust—even under messy, high-stakes conditions. We focus on the infrastructure and tooling behind what we call “data agents”: systems that reason, interact, and execute against live business data.