Delivery Project Assistant
Redesigning a fragmented AI experience as a single, agentic workflow
Role
Lead Product Designer
Client
AlphaSights
Launched
May 2026
Context
Delivery Project Assistant is an agentic tool that supports associates in managing their fragmented workflow end-to-end. The agent automates the tedious parts, like writing profiles and and tracking project statuses, so associates can focus on what matters: delivering quality experts quickly for clients.
My Role
I got to lead the design effort on creating a new modality to house moonshot features, shaping a brand new framework for our org's approach to AI tooling.
Self-initiated and self-driven
This wasn't a planned initiative. I originated it. While auditing our internal platform, I noticed that our AI initiatives were built in isolation, treated as individual features on different pages, and owned by different teams.
I made the case to build one modality to serve multiple needs through independent research, benchmarking, and early design explorations.
I leveraged Cursor to build working prototypes. Developers took my code and built it directly, making something that had felt unachievable suddenly within reach. This led to stakeholder buy-in, and a roadmap followed.
Reconciling the current solutions
To design it, I had to reconcile with how associates work on our internal platform. I adapted their nuanced mental models into a new modality that still maintained consistent patterns users depended on.
Expanding the design system
I extended AlphaSights Design Systems into a new context. I introduced an extension of our design system to include AI patterns, establishing new components and laying the groundwork for cross-domain alignment.
What we shipped
Our proof of concept was presented at our company-wide Quarterly Update in Q1 2026 as well as our domain offsite in Q4 2025.
Core features are planned to be globally released in May 2026.
AI Project Launch
AI First Deliverable
AI Assisted Outreach
Recorded and Transcribed Vetting Calls
Reflections
Through this project, I found joy in navigating ambiguity, setting precedent, and driving product direction.
Designing for associates' nuanced workflows in an emerging modality meant there was no established playbook for what "right" looked like. It deepened my understanding of AI interaction patterns and sharpened my instincts for a design principles that are still being defined in real time.
It also made me AI-native in how I work — not just in what I shipped, but in how I research, prototype, and collaborate.