Agentic Experience Design
This case study showcases the design work I led and contributed to (2025-2026), for producing net-new, redesigned Xfinity member-facing experiences leveraging LLM technology and agentic workflows. This work is set to impact millions of Comcast Xfinity customers by the end of 2026.
Minimum viable experience for Proactive Outage Communications demonstrating 3 capabilities - Contextually aware text response, Natural language utterance & turn-taking, and Rich Business Messaging format.
Minimum viable Xfinity Assistant Web chat experience for repair and troubleshooting agent.
Overview
This page showcases 2 separate projects involving Xfinity Assistant (XA) agents, one designed for an SMS-based experience and the other for a web chat experience. While I led the experience for the Outage Agent (pending launch), I contributed to the Repair and Troubleshooting agent over time and led the creation of a demo for senior executives. Initial UI/UX explorations set the tone for this XA repair and troubleshooting agent design and development, which is now being piloted with 1% of employees and is pending a full launch to millions of XA customers by the end of this year (2026).
1. Agentic Outage Experience
I led the design of the prompt framework and proof-of-concept demo for the 2026 next-generation outage solution, pioneering agentic and contextually intelligent messaging. The demo demonstrated three new features during a complex event where a planned network maintenance window overlapped with an unplanned severe-weather incident.
Xfinity climbed from last place in 2022 to first place in timely, reliable proactive outage communications in 2025 — J.D. Power.
The Problem-
Today’s messaging experience for outages doesn’t consistently do the following:
A) Provide members with a unified experience across outage types (planned/ unplanned) or channels (proactive/reactive).
B) Deliver critical updates in a timely manner.
C) Assist with outage-related member inquiries.
This can result in a member experience that feels robotic and increases members' effort when contacting support during an already frustrating event. This reduces members' confidence in Xfinity’s automated service comms.
Goal-
Deliver a unified, proactive, effortless messaging experience that drives intelligent conversations with members impacted by 3 or more planned or unplanned outages within a month.
My Role-
Designed reusable prompt framework, conversational script, agentic interaction, and demo.
Existing outage notifications bombard members (customers) with status updates almost daily or every few days as the problem code/job type changes from planned to unplanned outage.
Initial concept exploration
The minimum viable experience demonstrated three capabilities: contextually aware responses, natural-language utterances with turn-taking, and rich business messaging.
Ingredients of the reusable prompt framework
The framework brought together member and outage context, journey state, business rules, response guidance, escalation behavior and channel constraints so the agent could adapt while remaining predictable and safe.
Next steps
The demo was ready and the next step was to partner with an engineer and a solution architect to stitch all the pieces of the prompt together, gauge the technical feasibility of the MVP design and hook up the system prompt architecture with tools APIs, MCP etc. to launch Phase 1 of MVP showcasing the 3 capabilities demonstrated above. Per my last communication with the PM, I know that this experience was set to target 50% of Xfinity residential customers by the end of 2026. But the timeline for this experience is contingent on the full rollout of RCS messaging to all Xfinity members.
2. Repair and Troubleshooting Agent
Working backward from the ideal-state visual proof of concept, I contributed to fine-tuning the repair and troubleshooting agent in GitHub architecture linked to Google ADK (Agent Development Kit), which powered Xfinity's customer-facing experiences. This work spanned over four quarters discontinuously. I partnered with various cross-org. and cross-functional stakeholders.
Problem: High-effort customers (~482K daily) faced significant friction in resolving issues, resulting in 10.6x more time spent in support, higher costs, and increased churn risk. This presented a clear opportunity for GenAI optimization.
Goal: 1- Enhance current troubleshooting from a guided, step-by-step, high-effort experience with moment-in-time telemetry to a conversational experience with real-time, historical, and contextual telemetry. 2- Reduce repeats, as in reduce the number of times customers have to call in or reach out about the same issue multiple times through different channels.
My Role: Design the agentic interaction and prototype a production-ready, telemetry-driven experience for repair and troubleshooting agent.
Vision and ideal-state experience
The initial exploration established the experience vision before LLM resources and technology became available for use.
Targeted areas of improvement
The prompt framework focused on stronger context retention, more effective use of telemetry, clearer response structure, reduced repetition, natural turn-taking and graceful escalation when automation could not resolve the issue.
