DeveloperWeek 2025

Saahil Jain, You.com, Head of AI
Bryan McCann, You.com, Co-Founder & CTO

Description: In this session, You.com CTO Bryan McCann and Head of AI Saahil Jain will discuss various ways businesses can use AI to power new experiences across use cases, leveraging our experience in information retrieval and generative AI since we first launched in 2020. As the AI space rapidly evolves, You.com is constantly thinking about how to stay ahead of the curve and provide answers to the increasingly complicated questions that businesses face. We will share best practices and tips on how to use the latest advances in AI effectively.

Lukasz Pater, CERN, Software Engineer

As JavaScript continues to dominate web development, effective package management has become essential for navigating the increasing complexity of modern applications. This talk dives into the mechanics of three leading JavaScript package managers—npm, Yarn, and pnpm—shedding light on their distinct approaches to dependency management, performance, and disk space optimisation. Attendees will gain a clear understanding of how each tool handles node_modules, along with their advantages, trade-offs, and guidance on choosing the right package manager for any project.

We’ll also take an in-depth look at package.json, the core of every JavaScript project. Beyond just listing dependencies, this file offers powerful features like scripts, peer dependencies, versioning strategies, and more to help streamline project management. In addition to package management, the session will explore the JavaScript module ecosystem, its inherent challenges, and how package managers simplify—or complicate—the process of module loading.

Finally, we’ll cover advanced topics like zero-install setups, workspaces, polyglot tools, the impact of package managers on DevOps workflows, along with emerging players like Bun, Deno, and Yarn Berry.

Eleftherios Chrysochoidis, Chubb, Lead API Software Engineer

GitHub Actions offer robust and free CI/CD capabilities to streamline your development process. In this presentation, we'll dive into the core components of GitHub Actions and explore how they can be set to work for you by demonstrating various simple and advanced use cases!

GitHub Actions is an ideal tool for any project hosted on GitHub. It provides CI/CD features out of the box and is accessible to everyone without the need for extra infrastructure or setup. The declarative way of creating Actions makes their usage really easy, and its remarkable free layer provides sufficient limits even for big projects.

Especially for Open Source projects, where usually there is more than a single contributor, the need for CI/CD is huge. What will happen if you own an open source project and someone creates a Pull Request with really nice features but adds Security risks in the project (e.g. due to deprecated dependencies) or if they break some of the existing functionality? Would you be able to check all of them on your own? Maybe yes, but it would take much more time compared to having them all checked automatically by utilizing GitHub Actions.

Join me to explore how to make GitHub Actions run faster by utilizing caching mechanisms, build Reports for Testing, Coverage and Security issues and integrate them with any open Pull Request in the community, to have an automated evaluation of the suggested PR and get notifications for them.

Manan Bedi, Paytm, Business Analyst, GenAI

AI is the most hyped topic in the current tech industry, especially with the advent of LLMs and GenAI. But never to forget, JavaScript is still one of the most used and popular programming languages among Software Engineers. TensorflowJS is one such framework through which we can build AI Applications using JavaScript. The session will explore using TensorflowJS to build amazing AI-enabled applications using a pre-trained CocoSSD Model. We will also dive into creating Neural Networks from Scratch using the framework.
We will understand the Client Side and Server Side powers both and how TensorflowJS utilizes both.
Additionally, we will also do a hands-on to build a Smart WebCam to identify objects with high accuracy along with the frontend built using HTML and CSS.
Hence, it would be a great opportunity to understand and implement AI in our projects using our favorite language: JavaScript.

Jason Kelly, Fremont AI, CEO

POCs and prototypes are exciting, but productionizing and scaling Generative AI applications is hard. How do you effectively monitor compliance, privacy, and safety? How do you support multiple models and use cases while avoiding vendor lock-in? And how do you help teams build their apps without reinventing the wheel or creating security vulnerabilities?

Whether your developers are wrapping third-party GenAI APIs or self-hosting a complex stack with RAG frameworks and fine-tuned LLMs, the landscape is filled with tools and frameworks that promise solutions. In this vendor-agnostic session, you'll get a detailed look at GenAIOps principles and tools. With real-world examples and sample code, you’ll learn how to navigate these challenges and leverage Generative AI with confidence.

We'll cover:

- Examples of GenAI pitfalls and business risk they create
- Components of a GenAIOps stack: Monitoring, Guardrails, API Management, Security, Developer Experience
- How GenAI integrates with the standard SDLC
- Examples from interviews with real-world teams in industry
- How-To: Deployable sample code for a cloud environment showing these guidelines in action [GitHub link to be included]

Nimisha Sharath, Uber, Senior Product Manager

Recall those AI-infused products that stumbled in the wild? They caused rollbacks, discontinuations, and hefty financial losses for companies. Nimisha, having crafted and launched numerous ML/AI products across organizations, shares her battle-tested blueprint to launch successful AI products that scale.

From product discovery to legal considerations, metrics, AB testing, GTM, and model quality maintenance, she guides you toward creating AI products that truly endure.

Andrew Schall, ServiceNow, UX Research Leader & Strategist

Allowing organizational fortunetellers to create a future state vision of a journey map is risky because their projections are often based on speculation, intuition, and vague aspirations rather than concrete data and a deep understanding of the current state.

The journey from the current state to the optimal future state can be envisioned as constructing a long metaphorical bridge. This bridge is supported by various foundational pillars, each representing a critical aspect of the progression. These pillars provide the necessary support for incremental solutions to realize the optimal future state.

In this session, you will delve into envisioning an ideal user experience by creating a future state journey map. You will learn how to ground your future state journey map in current realities based on data and user research, ensuring that solutions are practical, actionable, and directly address identified pain points. By following this structured, evidence-based methodology, organizations can successfully navigate the journey from the current state to the future state, building a robust bridge to support ongoing progress, innovation, and adaptation.