Opportunity with Sesame
Thank you for the opportunity to share my work with you and Sesame, Anes!
Above, you’ll find a link to an overview of my portfolio and my resume. Below, I’ve provided a deep dive into two representative works below. Though I can't share my current Meta projects for legal reasons, these selections offer the best 'behind-the-scenes' look at how I tackle the types of agentic and natural interaction challenges Sesame is currently solving.
Questions from Sesame
Why are you looking to leave your current position?
1
I’m currently working on interaction design for AI-driven systems across a multi-device ecosystem, which has been a great experience.
At this stage, I’m looking to increase my level of ownership. Much of the high-level direction for the agentic system I’m working on is set at a broader organizational level, and my role has been more focused on execution within that framework.
I’m interested in roles where I can contribute more directly to defining the interaction model itself - helping shape how the system behaves from the ground up.
Why are you interested in Sesame?
2
Getting someone to talk to AI is easy. Getting them to stay, engage, and come back is the real problem.
What stands out to me about Sesame is that you’re tackling a harder version of the voice problem.
Most systems today are still primarily transactional - ask something, get a response. What you’re building seems more focused on making interactions feel natural and continuous, which is a very different challenge, and one that gives your system the opportunity to learn and become personalized.
That shift - toward timing, tone, and conversational flow - is something I’ve been thinking about a lot in my recent work with AI systems, so it feels like a strong fit.
How would you approach Sesame’s problem and apply your expertise?
3
Having spoken with Miles and Maya at length, what stood out to me is that Sesame is one of the first voice agents that actually encourages continued interaction. That feels like the core challenge: moving from single-turn novelty to sustained, multi-turn engagement.
If the experience breaks - latency, awkward timing, or responses that don’t feel natural - people disengage quickly. And when that happens, the system never really learns the user. So this isn’t just a UX problem, it directly impacts how intelligent the product can become over time.
I’d approach this as a system, but also make sure it translates into clear product decisions, interfaces, and measurable outcomes.
1. Earn continued engagement through interaction quality
I’d start with timing, responsiveness, and interruptibility - making sure the system feels natural enough that people actually want to stay in the conversation.
2. Design for fluid, multi-turn interaction
Beyond conversation quality, I’d map how interactions evolve over time - handling interruptions, shifting intent, and follow-ups without breaking flow.
3. Build an output orchestration layer for agentic behavior
I’d focus on how the system interprets signals across moments and decides what to say, what to do, and when to do it. This is where it starts to feel like an agent instead of a chatbot.
4. Translate behavior into product surfaces (UI, flows, and states)
Even in a voice-first system, there are critical supporting surfaces - onboarding, feedback states, history, controls, and edge cases. I’d define flows & lightweight UI to support the voice experience without getting in the way.
5. Validate through qualitative and quantitative research
This is a space where intuition isn’t enough. I’d pair rapid prototyping with qualitative testing to understand how interactions feel, and then use quantitative methods to measure engagement - drop-off rates, multi-turn depth, return usage, and feature adoption.
This is closely aligned with the work I’ve been doing across AI and wearable systems: designing not just the interaction itself, but how it holds up over time, how it’s supported through product surfaces, and how it’s validated through real user behavior.
What is a negative signal in your current position that you wouldn’t like to see in your next?
4
One thing I’ve learned is how important it is for interaction design to be involved early, especially when the product direction is still forming.
In larger environments, interaction work can sometimes come in later, after key decisions are already made. That tends to limit the impact of the work.
I’m looking for a team where interaction design is part of shaping the product from the beginning, particularly on problems that don’t have an established playbook.