Software Development Manager For Amazon Ads AgentWe're building the AI-powered agent experience for Amazon Ads the chat-first interface that will change how hundreds of thousands of advertisers manage campaigns, analyze performance, and optimize strategy on Amazon. We need an SDM to lead this from the front.

Amazon Advertising helps advertisers reach and engage audiences across our digital properties, premium third-party sites, streaming TV, and millions of devices. The Advertiser Experience Frameworks team owns the experience layer of the Ads Agent the chat interface, advertiser-facing communications, feedback systems, and the frameworks and services that enable dozens of partner teams to integrate agent capabilities into their pages across the Amazon Ads Console.

We are looking for a Software Development Manager to lead a team building these experiences and the platform that powers them. This is not a purely technical execution role. You'll need strong product instincts the ability to look at an advertiser workflow, understand where friction exists, and form a point of view on how agent-driven experiences should work before a PRD lands on your desk. You'll need to either bring knowledge of the advertising business or demonstrate the curiosity and learning speed to develop it quickly, because the quality of your technical decisions depends on understanding how advertisers actually think about campaigns, budgets, targeting, and performance measurement.

The role spans the full web stack from front-end chat interfaces and UI frameworks to middleware services, APIs, event architectures, and the operational infrastructure required to run a global, high-availability website. You'll drive the integration of agent-driven workflows into multiple surfaces across the Ads Console, while also building and maintaining the framework layer that partner teams depend on to bring these capabilities into their own console pages.

This is a horizontal team role. Dozens of teams across Amazon Ads will integrate with your frameworks and services. The right candidate thrives in this model you're comfortable setting technical direction that other teams build on, skilled at negotiating API contracts and integration patterns across organizational boundaries, and experienced at maintaining platform quality and reliability even as adoption scales rapidly. You know how to balance speed and quality when both matter equally: moving fast enough to stay ahead of a rapidly evolving space while maintaining the bar that partner teams and advertisers depend on.

The right candidate has built and operated web systems at scale across multiple layers of the stack, has recent hands-on experience building AI-powered products (particularly with LLM-based or agent-oriented systems), and brings product thinking to every technical decision. You don't need a decade of AI experience, but you do need real, demonstrable work with these technologies, a strong point of view on how they change both software development practices and user experiences, and either advertising domain knowledge or a track record of ramping quickly in complex business domains.

Key Job ResponsibilitiesLead and grow a high-performing engineering team delivering the Ads Agent chat experience, supporting services, and the agent integration framework that dozens of partner teams use to bring agent capabilities into their Ads Console pages.

Drive technical vision and architecture across the full web stack front-end frameworks, middleware services, APIs, and event architectures making decisions on skills-based versus agent-based approaches, integration contracts, and patterns that scale across the organization.

Bring strong product thinking to technical decisions develop deep understanding of advertiser workflows, pain points, and goals so that architecture and experience decisions are grounded in how advertisers actually use the product, not just what's technically elegant.

Operate as the technical leader of a horizontal platform

team: set direction that other teams build on, negotiate integration patterns across organizational boundaries, and maintain framework quality and reliability as adoption grows.

Balance speed and quality as a core discipline move fast enough to lead in a rapidly evolving space while maintaining the reliability, performance, and operational bar that partner teams and hundreds of thousands of advertisers depend on.

Partner with Product and UX as a true thought partner not just translating requirements into technical solutions, but actively shaping the advertiser experience vision based on technical possibilities, business context, and user behavior insights.

Partner with Applied Science teams to define evaluation methodologies and tooling that ensure agent response quality meets a high bar before reaching advertisers.

Establish engineering best practices for AI-native development including design review mechanisms, technical specification standards, and coding patterns tailored to agent-driven systems.

Drive operational excellence for a global website owning availability, latency, monitoring, alarming, and incident response across the services and frameworks your team delivers.

Mentor and develop engineers through a significant technical transition, building team capability in AI-powered development and product-minded engineering while maintaining delivery velocity.

A day in the life

You start your day reviewing operational metrics and system health across the full stack front-end performance, middleware latency, API error rates, and event pipeline throughput. When something looks off, you dig in with your team to understand whether it's a rendering issue, a service degradation, an event delivery failure, or an agent quality regression and you drive the right team to resolution.

Mid-morning, you're in a technical design review. Your team is proposing the architecture for a new framework capability that will allow partner teams to integrate the chat experience into campaign management pages. You ask sharp questions that blend technical depth with product thinking: How does this handle advertisers with 500

• campaigns? When an advertiser asks the agent to adjust bids across a portfolio, what does the interaction flow look like and does it match how advertisers actually think about bid management? What's the API contract partner teams will code against? How do we version this without breaking existing integrations? What's the fallback when the agent can't fulfill a request and does that fallback feel natural to an advertiser mid-workflow? Your depth across the web stack, your understanding of advertiser needs, and your experience with AI-powered systems make these reviews substantive, not ceremonial.

Your key stakeholders include:

Dozens of partner engineering teams across Amazon Ads who integrate your chat framework and services into their console pages this is your most critical relationship surface and a defining characteristic of this role

Product Managers defining the advertiser experience vision you're a thought partner here, not just a recipient of requirements

UX Designers crafting the chat and console interfaces

Orchestration and reasoning teams building the layers beneath your experience and service layer

Applied Science teams working on agent quality, evaluation, and model behavior

After lunch, you have back-to-back conversations with two partner teams. One is integrating the chat framework into their reporting page and has questions about event handling and state management. The other is requesting a new API capability to support a custom workflow for sponsored brands advertisers. You help your team evaluate the request: Is this a one-off need or something that belongs in the platform? Does it reflect a pattern we'll see from other advertiser segments? How do we deliver it without slowing down the roadmap? This is the daily rhythm of a horizontal team managing dozens of integration points while protecting your team's ability to move forward on your own priorities.

Later, you join a product review where the team is looking at Voice of Customer data from advertisers using the agent experience. You notice that small and medium advertisers are using the agent differently than large advertisers with managed accounts they're asking more exploratory questions rather than executing specific actions. You raise this with the product manager and propose an experiment to test whether the agent should adapt its interaction style based on advertiser segment. This is the kind of product thinking that distinguishes great SDMs on this team you don't wait for someone to hand you a spec; you see patterns in the data and form a point of view.

Late afternoon, you join a cross-team architecture discussion with partners in a different time zone. The topic: whether a new advertising capability should be built as a skill within the existing agent or warrants a separate agent. You bring a framework your team developed for making these decisions and drive alignment.

You close the day in sprint planning, helping your team prioritize a new partner integration against technical debt from a prototype that shipped fast and needs hardening. You push for a concrete plan that delivers value this sprint while paying down the debt over the next two because in a horizontal team, the cost of accumulated shortcuts compounds with every new partner that builds on your platform.

About the team

The Advertiser Experience Frameworks team is approximately 35 engineers building the experience layer and platform services for the Amazon Ads Agent and the Amazon Ads Console. We own the chat interface, advertiser-facing communications, feedback collection, Voice of Customer programs, and the frameworks and services that enable dozens of partner teams to integrate agent capabilities into their pages across the Ads Console. Partner teams own orchestration and reasoning we own what advertisers see and interact with, and we provide the platform that makes integration seamless.

This is a horizontal team by design. Our frameworks and services are integration points for teams across the Ads organization. That means our technical decisions have outsized impact the patterns we set, the APIs