Name: TechLead Conf London 2025: Adopting AI in Orgs Edition
Start: 2025-11-28T08:00:00.000Z
End: 2025-11-28T17:00:00.000Z

Taz Singh

Daria Shabala

Georgios Giatsidis

Antanina Vertsinskaya

Theodoros Giatsidis

Joon Park

Beth Swingler

Henrik Tomson

Oury DIALLO

Dominik Kościelak

Ivan Almasi

FLOR URBINATI

Milos Adzic

Dom Sipowicz

Carlos Parreno

Kim Diep

Spiros Diochnos

Archana Gupta

Robbie Thandi

Julián Parrilla Calvo

Karina Bunta

Andrew Murphy

Esta masterclass introduce cómo usar LLMs de código abierto on-premise para un mejor control de datos, privacidad y eficiencia de costos. Destaca Ollama para la gestión simplificada de LLM (descarga, ejecución de modelos offline) y FastAPI para un alojamiento local eficiente, permitiendo un desarrollo rápido de API e integración sin problemas. Este enfoque combinado ofrece una seguridad superior, cumplimiento normativo y personalización en comparación con las soluciones basadas en la nube.

On-Premise Open-Source LLMs con Ollama & FastAPI

Ko Brendan hosts workshop on setting up AI models and leveraging Olyama framework. Training AI models is intensive and involves specific datasets. Fine-tuned AI models grow with additional information. AI models process annotation labels to find patterns. Fine-tuned models provide accurate responses for chat interactions. Setting up AI environment for faster response times. Testing AI models with job descriptions and analyzing applicant profiles. Exploring AI model enhancement capabilities for automating processes. Configuring model interaction and switching on Olyama. Utilizing database for specific company data retrieval. Development stack includes Next.js, React, and Tailwind for user-friendly features.

This workshop introduces how to use open-source LLMs on-premise for enhanced data control, privacy, and cost efficiency. It highlights Ollama for simplified LLM management (downloading, running models offline) and FastAPI for efficient local hosting, enabling fast API development and seamless integration. This combined approach offers superior security, regulatory compliance, and customization compared to cloud-based solutions.

On-Premise Open-Source LLMs with Ollama & FastAPI

Brendon Co

Los líderes técnicos enfrentan el desafío de equilibrar la innovación con la gobernanza al implementar asistentes de IA en equipos. Esta sesión ofrece una visión táctica sobre la operacionalización de la IA en el desarrollo de software, cubriendo métricas de éxito, telemetría, bucles de retroalimentación y salvaguardas empresariales. Basado en el marco GitHub Well-Architected y ampliado a herramientas de IA de propósito general, exploraremos cómo crear visibilidad en el uso, abordar riesgos de seguridad y cumplimiento, y asegurar que el desarrollo potenciado por IA no solo ocurra, sino que escale con intención y confianza. Imprescindible para aquellos que construyen las plataformas y políticas de las organizaciones de ingeniería del mañana.

De Experimento a Empresa: Escalando Asistentes de Codificación de IA a Través de Equipos de Ingeniería

Sharing insights on scaling AI coding assistance & adopting AI native dev tools. Discussing the need for a shift from viewing adoption as a tech problem to a change management challenge. Exploring a tactical framework with three key phases: onboard, adopt, succeed. Emphasizing the importance of onboarding developers, ensuring access to licenses, and setting clear policies to avoid 'shadow AI'. Highlighting the necessity of human infrastructure for successful AI dev tool implementation, including AI advocates, communities of practice, and executive sponsorship. Adoption strategies require a focus on interconnected metrics and all pillars simultaneously. Leadership checklist for post-conference actions involves measuring metrics, distinguishing between strategic and tactical points, and staying connected through platforms like WhatsApp and LinkedIn.

Technical leaders face the challenge of balancing innovation with governance when rolling out AI assistants across teams. This session offers a tactical view on operationalizing AI in software development - covering success metrics, telemetry, feedback loops, and enterprise guardrails. Based on the GitHub Well-Architected framework and broadened to general-purpose AI tools, we’ll explore how to create visibility into usage, address security and compliance risks, and ensure that AI-powered development doesn’t just happen - it scales with intent and trust. A must-attend for those building the platforms and policies of tomorrow’s engineering organizations.

From Experiment to Enterprise: Scaling AI Coding Assistants Across Engineering Teams

Maxim Salnikov

A medida que la inteligencia artificial evoluciona de la simple automatización de tareas a agentes autónomos capaces de tomar decisiones independientes y ejecutar tareas, las empresas enfrentan nuevos desafíos de gestión de identidad y acceso (IAM). En este masterclass interactivo, los participantes explorarán y aplicarán estrategias para asegurar agentes de IA mientras consumen servicios empresariales en nombre de los usuarios o por su cuenta.A través de ejercicios prácticos, escenarios del mundo real y un tutorial técnico guiado, aprenderás cómo los agentes de IA pueden:- Ser tratados como entidades gestionadas y de primera clase dentro de los sistemas empresariales- Autenticarse y obtener autorización de manera independiente o en nombre de los usuarios- Aprovechar y extender estándares como OAuth 2.0 y OpenID ConnectAl final del masterclass, los asistentes tendrán habilidades prácticas y un marco para asegurar agentes de IA mientras preservan sus capacidades autónomas, asegurando la adopción empresarial sin comprometer la seguridad o el cumplimiento.

IAM para la Era Agente: Guía para Construir Agentes de IA Seguros

Welcome to a session focusing on AI agents, IAM, and demo setup by Godwin Namila from WSO2. Understand the role of agents in decision-making and task performance, and their significance for enterprises. Learn about the Model Context Protocol enabling standardized communication for agents. Explore securing AI model invocation with AI gateways and MCP server protection. Delve into the hotel system demo showcasing trip planning, booking, and agent functionalities. Gain insights into agent operations, IAM implementations, and AI access management standards.

As artificial intelligence evolves from simple task automation to autonomous agents capable of independent decision-making and task execution, enterprises face new identity and access management (IAM) challenges. In this interactive workshop, participants will explore and apply strategies to secure AI agents as they consume business services on-behalf-of users or own their own.Through hands-on exercises, real-world scenarios, and a guided technical tutorial, you’ll learn how AI agents can:- Be treated as managed, first-class entities within the business systems- Authenticate and obtain authorisation independently or on behalf of users- Leverage and extend standards such as OAuth 2.0 and OpenID ConnectBy the end of the workshop, attendees will have actionable skills and a framework for securing AI agents while preserving their autonomous capabilities, ensuring enterprise adoption without compromising security or compliance.

IAM for Agentic Era: Guide to Build Secure AI Agents

Godwin Amila

Las entrevistas técnicas están luchando por mantenerse al día con la realidad de la ingeniería de software moderna en una época de agentes de IA. Los LLMs y sus derivados ahora pueden resolver fácilmente evaluaciones para llevar a casa, y los formatos tradicionales de DSA recompensan la memorización sobre las habilidades del mundo real. Entonces, ¿cómo podemos evaluar de manera precisa y justa las habilidades técnicas de los ingenieros como parte de un proceso de contratación estandarizado? A finales de 2024, comencé un rediseño del proceso de entrevistas técnicas en HelloBetter, una empresa de tecnología de salud mental en Berlín. El objetivo era crear un formato que realmente reflejara el trabajo diario de ingeniería, al mismo tiempo que fuera consciente del tiempo, inclusivo y aceptara la presencia de IA en los flujos de trabajo de los desarrolladores. En el proceso, creamos "El Método McDougall": un marco nuevo y emergente para mejores procesos de entrevista que está obligando rápidamente a los líderes técnicos a replantearse sus estrategias de mitigación de IA.

Transformando tus Entrevistas Técnicas en la Era de los Agentes de IA

Anna McDougal, director of engineering operations at Blinkist, presents the McDougal method for technical interviews aiming to improve their effectiveness by leveraging emerging technologies. The challenges in technical interviews involve the inadequacy of measuring the right skills through traditional methods like DSA problems and take-home tests. The speaker questions the necessity of perfection in engineering hires and emphasizes problem-solving skills over perfection, especially with AI tools. Hiring criteria are evolving with a focus on attitude alignment and cultural addition rather than just fit. Maximizing value in hiring involves seeking candidates with aligned goals and the necessary toolkit for the job. Ensuring technical interviews reflect actual job conditions and tool usage is crucial, and encouraging experimentation and feedback on proposed methods is essential for improvement.

Technical interviews are struggling to keep pace with the reality of modern software engineering in a time of AI agents. LLMs and their offshoots can now easily solve take-home assessments, and traditional DSA formats reward memorisation over real-world skills. So how can we accurately and fairly assess the technical abilities of engineers as part of a standardised hiring process? In late 2024, I began a redesign of the technical interview process at HelloBetter, a mental health tech company in Berlin. The aim was to create a format that actually reflects day-to-day engineering work, while also being time-conscious, inclusive, and accepting of the presence of AI in developer workflows. In the process, we created "The McDougall Method": a new and emerging framework for better interview processes which is quickly forcing technical leaders to rethink their AI mitigation strategies.

Transforming Your Tech Interviews in the AI Agent Age

Anna J McDougall

Gregor compartirá tanto las tendencias positivas como negativas que está observando, hablando con muchos líderes de ingeniería en toda la industria. También compartirá cómo está evolucionando el rol de los líderes de ingeniería y consejos sobre qué hacer para ser un líder de ingeniería exitoso en la era de la IA.

Cómo la IA está impactando el liderazgo en ingeniería

The speaker discusses the impact of AI on engineering leadership and shares personal journey and goals. Sensationalistic AI predictions induce fear of missing out among engineering leaders, leading to negative perceptions. Unrealistic AI expectations affect motivation and industry sentiment. Strategies for effective AI adoption include starting small, prioritizing culture over tools, and staying informed. Engineering roles are evolving towards a convergence, focusing on human-related skills and problem-solving in the AI era. Balancing code writing and reviewing, supporting junior colleagues, and navigating career progression are crucial for engineering leadership development.

Gregor will be sharing both positive and negative trends that he's seeing, speaking with many engineering leaders across the industry. He'll also share how the role of engineering leaders is evolving, and tips on what to do to be a successful engineering leader in the age of AI.

How AI is Impacting Engineering Leadership

Gregor Ojstersek

A medida que la IA transforma el desarrollo de software, los equipos enfrentan una elección crítica: liderar el cambio o arriesgarse a quedarse atrás. En Typeform, estamos fomentando proactivamente el uso de herramientas de IA y ayudando a todos a mejorar su competencia en IA.Esto significa muchas cosas, incluyendo la flexibilidad para probar diferentes herramientas, democratizar el acceso a (pero controlar nuestro gasto en) LLMs, y asegurarnos de que estamos contratando a la próxima generación de ingenieros que tienen la IA como parte de su ADN.En esta charla, proporcionaré algunos ejemplos prácticos de cómo estamos logrando estas cosas en Typeform, así como cómo estamos mejorando la forma en que colaboramos con personas fuera de la ingeniería utilizando herramientas de IA.Compartiré ejemplos reales usando AWS Bedrock y servidores MCP, además de lecciones duramente ganadas de nuestro viaje. Si tu equipo está en el camino hacia la transformación de la IA, esta charla te dará algunas estrategias prácticas para hacerla un éxito.

Democratización de la IA en Ingeniería: Lecciones del Viaje de Typeform

Andy introduces himself at Typeform, discussing AI democratization and his interests in Star Trek and Emacs. The conversation delves into the hype and slow adoption of AI tools at Typeform, emphasizing the journey of democratizing AI access. The use of AWS Bedrock for accessing LLMs and the challenges of simplifying access are explored. Implementing Bedrock token generation, AWS convenience, and tool federation at Typeform are discussed. The talk also covers the impact of AI on hiring, evolving AI integration, responsible AI usage, assessing AI proficiency, internal productivity assessment, and knowledge sharing in tool adoption.

As AI transforms software development, teams face a critical choice: lead the change, or risk being left behind! At Typeform, we're proactively encouraging the use of AI tools, and helping everyone level-up their AI proficiency.This means lots of things including the flexibility to try different tools, democratising access to (but controlling our expenditure on) LLMs, and ensuring we're hiring the next generation of engineers who have AI as part of their DNA.In this talk, I'll provide some practical examples of how we're achieving these things at Typeform, as well as how we're enhancing the way we collaborate with people outside of engineering using AI tools.I'll share real examples using AWS Bedrock and MCP servers, plus hard-won lessons from our journey. If your team is on the path to AI transformation, this talk will give you some practical strategies for making it a success.

Democratising AI in Engineering: Lessons from Typeform's Journey

Andy Kuszyk

Nuestro mundo funciona con TI construida en los años 80 y 90 por ingenieros que ahora se están jubilando. Esto se está volviendo muy preocupante para las corporaciones que tienen sistemas críticos funcionando en código legado que solo un puñado de personas aún entiende.Ahí es donde llega GenAI. Con sus habilidades de transpilación casi mágicas, es la innovación que se necesitaba para iniciar las modernizaciones de legado que se habían pospuesto durante demasiado tiempo.Habiendo trabajado en modernizaciones críticas durante más de 15 años, aproveché la oportunidad para experimentar cómo los LLMs pueden acelerar proyectos de migración complejos.Compartiré la experiencia concreta que hemos acumulado en una amplia gama de stacks (Swift-&gt;React-Native, Eclipse RCP -&gt; Spring Boot, Java 1.6 Spring -&gt; Java 21 Spring Boot, PHP ZF1 -&gt; Symfony 7), cubriendo tanto las buenas sorpresas como las limitaciones que descubrimos, y compartiré nuestro manual actual para aprovechar al máximo la AI al migrar un sistema legado.

Modernización de Legado Acelerada por AI

Discussing the importance of adopting AI in legacy modernizations and the risks associated with outdated systems, challenges in legacy ecology, reliability rupture, and migration mutiny. Exploring Lean Tech at Theodos for maintaining agility and customer value, improving system architecture through visualizing endpoints and high-quality coding. Optimizing accuracy with agent coders, industrializing migration with LeanTech principles for 10x acceleration. Leveraging AI for product building, AI model evaluation, and the preference for Claude over GPT in migration.

Our world runs on IT built in the 80s and 90s by engineers who are now retiring. This is becoming very worrying for corporates who have critical systems running on legacy code that only a handful of people still understand.That's where GenAI arrives. With its almost magical transpilation abilities, it is the innovation that was needed to kickstart the legacy modernisations that had been postponed for too long.Having worked on critical modernisations for more than 15 years, I jumped on the opportunity to experiment with how LLMs can accelerate complex migration projects. I will share the concrete experience we have accumulated on a wide range of stacks (Swift-&gt;React-Native, Eclipse RCP -&gt; Spring Boot, Java 1.6 Spring -&gt; Java 21 Spring Boot, PHP ZF1 -&gt; Symfony 7), cover both the good surprises and the limitations that we uncovered, and share our current playbook to best leverage AI when migrating a legacy system.

AI-accelerated Legacy Modernisation

Fabrice Bernhard

A principios de 2025, un equipo de ingenieros hábiles recibió un desafío formidable: pasar de un conjunto de habilidades de ingeniería de software tradicional a lanzar una solución de seguros de IA agente de grado de producción en solo 16 semanas. Esta charla cuenta la historia de ese viaje. Se adentrará en el manual práctico y centrado en el humano desarrollado para una rápida mejora de habilidades en la era de la IA. El presentador, Tim Makin, compartirá cómo su equipo construyó no solo una "sala de máquinas agente" en AWS Bedrock y una suite de agentes colaboradores, sino también una cultura de aprendizaje intenso, continuo y seguridad psicológica. Los asistentes se irán con un marco replicable para transformar sus propios equipos en potencias nativas de IA.

Forjando la Sala de Máquinas Agente: Un Manual Centrado en el Humano

The speaker shares experiences in AI project production and the importance of leadership in increasing successful AI product launches. Challenges in implementing AI solutions include resource allocation and adapting to new technologies. The transition to Python for a project launch in September involved upskilling the team. Establishing clear expectations and career frameworks for engineers is crucial for success. Fostering growth through experimentation and innovation events enhances skills and collaboration. Effective communication, user experience focus, and continuous testing are key to project success in AI initiatives. Encouraging psychological safety and involving employees in AI adoption are important strategies for transformative journeys.

In early 2025, a team of skilled engineers was given a formidable challenge: go from a traditional software engineering skillset to launching a production-grade, agentic AI insurance solution in just 16 weeks. This talk tells the story of that journey. It will dive into the practical, human-centric playbook developed for rapid upskilling in the age of AI. The presenter, Tim Makin, will share how his team built not just an "agentic engine room" on AWS Bedrock and a suite of collaborator agents, but also a culture of intense, continuous learning and psychological safety. Attendees will leave with a replicable framework for transforming their own teams into AI-native powerhouses.

Forging the Agentic Engine Room: A Human-Centric Playbook

Tim Makin

Learn how to build a production-ready Model Context Protocol (MCP) server running on Cloudflare Workers. This hands-on workshop walks attendees through creating an MCP server, exposing tools, integrating external APIs, and deploying globally so AI assistants can call your tools in real time.Workshop outcome: By the end of the session, attendees will have a deployed MCP server and the knowledge to extend it so AI assistants can call their tools securely and efficiently.

Hands-on: Building MCP Servers with Cloudflare Workers

Confidence Okoghenun

Las demostraciones de agentes son brillantes; el uso diario es desordenado. En esta charla compartiré un puñado de historias cortas y honestas de flujos de trabajo de agentes con equipos reales: lo que intentamos, cómo reaccionó la gente, qué nos sorprendió y qué cambiamos al día siguiente. En lugar de teoría, escucharás momentos concretos (éxitos, fracasos, casi fallos) y los aprendizajes prácticos que surgieron. Te irás con principios claros y probados en batalla que abarcan el nuevo mundo al que todos nos estamos acercando.

Todos son ahora un gerente - Interfaces en el mundo de los agentes

Sam Kmezverk's talk covers the evolving tech landscape, emphasizing key principles for agents' implementation in tasks and decision-making processes. Addressing challenges in agent orchestration, debugging system issues, and cultivating user trust are crucial for successful interface development.

Agent demos are shiny; daily use is messy. In this talk I’ll share a handful of short, honest stories from shipping agent workflows with real teams—what we tried, how people reacted, what surprised us, and what we changed the next day. Rather than theory, you’ll hear concrete moments (wins, faceplants, near-misses) and the practical learnings that emerged. You’ll leave with clear and battle-tested principles that encompass the new world we are all approaching.

Everyone Is Now a Manager - Interfaces in the World of Agents

Sam Komesarook

Cuando cada empleado puede crear agentes de AI, la innovación deja de ser un departamento y comienza a ser una cultura. La verdadera innovación rara vez es el resultado de una sola chispa: es el producto de un entorno donde todos están equipados y empoderados para contribuir con ideas. En Semble, he liderado la adopción y personalización de Dust, convirtiéndolo en una plataforma interna de AI que empodera a cada empleado para crear, probar y desplegar agentes, sin necesidad de conocimientos técnicos.  Al reducir las barreras para la experimentación e implementación, hemos asegurado que la innovación no esté confinada a la élite técnica. En cambio, se convierte en una experiencia vivida y diaria para cada empleado. Basándome en resultados del mundo real en Semble, esta charla ofrecerá ideas prácticas para líderes tecnológicos que buscan desbloquear el potencial creativo de sus equipos. Aprenderás cómo pasar de la innovación de arriba hacia abajo a un modelo participativo, por qué la gestión del cambio interno es tan crítica como la excelencia técnica y cómo hacer de cada empleado un campeón de la transformación. Únete a mí para descubrir cómo empoderar a las personas en todos los niveles puede redefinir lo que es posible para tu producto, tu cultura y el futuro de tu empresa.  (Descargo de responsabilidad: este resumen fue escrito por uno de esos agentes mencionados.)

Innovación Sin Límites: El Plano para una Cultura Donde Cualquiera Puede Construir Agentes de AI

The Talk highlights the integration of Semble and AI, focusing on identifying use cases over technology for organizational efficiency. Claire's empathy-driven AI development showcases the importance of user empathy in software creation. Semble emphasizes deep understanding of technology for innovation. The adoption of technology at Semble is driven by simplicity, learning-by-doing, and internal champions. Safety, guidance, and experimentation are key for agent usage. Leveraging customer insights leads to innovative ideas, while secure data handling empowers innovation. Transitioning to production scale with AWS and cultivating technology adoption through gradual showcasing of benefits are essential processes at Semble.

When every employee can create AI agents, innovation stops being a department and starts being a culture. True innovation is rarely the result of a single spark – it’s the product of an environment where everyone is equipped and empowered to contribute ideas. At Semble, I’ve spearheaded the adoption and customisation of Dust, turning it into an internal AI platform that empowers every employee to create, test and deploy agents, no technical background required.  By lowering the barriers to experimentation and implementation, we’ve ensured that innovation is not confined to the technical elite. Instead, it becomes a lived, daily experience for every employee. Drawing from real-world results at Semble, this talk will offer actionable insights for tech leaders looking to unlock the creative potential of their teams. You'll learn how to shift from top-down innovation to a participatory model, why internal change management is as critical as technical excellence and how to make every employee a champion of transformation. Join me to discover how empowering people at every level can redefine what’s possible for your product, your culture and your company’s future.  (Disclosure: this synopsis was written by one of those said agents.)

Innovation Without Limits: The Blueprint for a Culture Where Anyone Can Build AI Agents

Mikael Landau

En Capital on Tap, el equipo de IA comenzó fuera de Ingeniería y ha tenido un tiempo polémico integrando el equipo en nuestra estructura de Ingeniería existente, con un mandato poco claro y presión para convertir prototipos en sistemas de producción. La mayoría de las organizaciones tienen prototipos; muchas menos tienen productos de IA seguros y confiables para el cliente. Esta charla es el manual de un practicante para la IA en organizaciones y la medición del impacto de la IA y el ROI en un contexto fintech regulado del Reino Unido, centrado en el modelo operativo, puertas, preparación del producto y medición posterior al lanzamiento vinculada a resultados operativos.

Envío de IA Bajo Restricciones: Construir, Comprar o Eliminar

Dave, Engineering Manager at Capital Untap, shares a story of AI failure and lessons learned from moving too fast in AI development within the company. The AI team focused on developing a chatbot for customer service, which became a flagship initiative under engineering governance. Issues with accuracy and complexity led to pausing the chatbot project and exploring third-party vendors. Capital OnTap reset AI processes by integrating AI engineers into product teams, emphasizing standard engineering over specialized AI teams.

At Capital on Tap, the AI team began outside Engineering, and has had a contentious time integrating the team into our existing Engineering structure - with an unclear mandate, pressure to turn prototypes into production systems. Most organisations have prototypes; far fewer have safe, reliable customer-facing AI products. This talk is a practitioner’s playbook for AI in organisations and measuring AI impact and ROI in a regulated UK fintech context—focused on operating model, gates, product-readiness, and post-launch measurement tied to operational outcomes.

Shipping AI Under Constraints: Build, Buy, or Kill

Dave Glencross

TechLead Conference

I'm attending TechLead Conference 2025 – get your free remote ticket and join me there

The Conference for Tech Leads, Staff Engineers, and Technical Eng Managers.TechLead Conf 2025 tackles two critical challenges facing technical leaders today: navigating AI adoption in organizations and reducing system complexity. Through real-world case studies from startups to Big Tech, senior engineers and tech leads will share practical insights from the trenches.Engage in discussion rooms, hallway track with experts, hands-on practical workshops, and tens of insightful talks.