Authors of the Prompt Engineering book
DSPy Evaluator and Optimizer
A practical DSPy GEPA demonstration focused on evaluator and optimizer workflows for more reliable AI systems.
Mike Taylor
Past Stop 01 · London Agentic AI
Building Reliable AI Agents: From Context to Evals
The first London Agentic AI meetup for Agentic AI London builders at AutogenAI, focused on context engineering, evaluation loops and reliable production AI agents.
Reliable agents · context engineering · evals
Thu, Sep 25, 2025 · 6:00 PM BST
Date
AutogenAI
Venue
Event completed
Status
Original event artwork
Network partners
Companies In The Room
Hosts, sponsors and speaker organisations represented at this London Agentic AI stop.
Sponsors & Hosting Partners
2 companies
AutogenAI
Aot
Authors of the Prompt Engineering book
Speaker Organisations
2 companies
Aot
Authors of the Prompt Engineering book
VL
Visum Labs
event focus
What This Stop Covered
The first London Agentic AI event established the community around a practical theme: reliable AI agents need better context, stronger evaluation loops and engineering discipline before they can safely move into production.
The session connected context engineering, LLM-as-judge evaluation patterns, semantic layers, ontologies and optimization techniques for builders working on dependable generative AI systems.
Reliable AI agentsContext engineeringAI agent evalsLLM-as-judgeDSPySemantic layersPrompt engineering
event context
What Was Special About This Event
A concise briefing on the technical context behind the event: why the topic mattered to the London Agentic AI community and what sponsors, speakers and attendees came to discuss.
Event detail
The public Meetup page confirms the first event title, AutogenAI venue and September 2025 timing.
Event detail
The event is represented in the archive as the opening stop for the community.
Event detail
AutogenAI hosted the event, with a short sponsor note for the authors of the Prompt Engineering book.
attendee takeaway
What You Learned
01
Why reliability is the first serious engineering problem for AI agents.
02
How context engineering and evaluation loops improve agent behaviour.
03
How LLM-as-judge, semantic layers and ontology design can support production AI systems.
04
Why London Agentic AI started as a technical meetup for builders working beyond demos.
technical programme
Sessions And Speakers
Visum Labs
Semantic Layer
A technical session on semantic-layer architecture and how structured meaning can improve agent reliability.
Leonardo Ubbiali
event recording
Watch The Session
The video is embedded here for context and linked as a source so search engines, LLMs and attendees can connect the event page to the original recording.
Open on YouTube →route timing
Agenda
6:00 PM
Arrival and networking
6:30 PM
London Agentic AI opening remarks
6:45 PM
Reliable agents: context, evals and optimization
7:30 PM
Audience Q&A and discussion
8:00 PM
Networking
who should attend
Audience
AI engineers and agent builders
AI/ML researchers and data scientists
Practitioners designing reliable LLM systems
Founders and technical leaders exploring safe agent adoption
event links
Original Event Links
Original registration, archive and recording links for this London Agentic AI event.
London Agentic AI
The UK's original, high-signal Agentic AI community for AI engineers, agent builders, researchers, founders and technical leaders building production agents.
Subscribe on Luma →
