A CLEAR ROLE BOUNDARY
The assistant shipped.
I made Home Security answerable.
I did not lead or design the Home Assistant. I led two designers in building a planned Home Security RAG layer—structuring product truth so BotBrain could return clearer, safer guidance.
Led the Home Security RAG source of truth.
Authored the core corpus and prompts, assigned sections across my two-person design team, and partnered with Sarat's squad on grounding, failure modes, and implementation requirements.
The team established the assistant surface, privacy disclosure, feedback controls, and embedded self-service.
Team-owned. I tested it, tried to break it, and advised in selected strategic areas.
My direct leadership: corpus, prompts, grounding, failure modes, and safe resolution paths.
Built by Sarat's squad. My team prepared the Home Security knowledge layer for ingestion.
PROOF OF AI PRODUCT CRAFT
I turned product truth into prompt-ready structure.
For Home Security, I led two designers in authoring prompts and grounding content for BotBrain—translating product rules into entities, context, constraints, failure modes, and safe next actions.
Home Security RAG source material
OPEN FULL RESOLUTION ↗
Technical truth, translated.
Define the ontology.
Normalize plans, devices, customer states, and product boundaries into a structure the system can retrieve.
ENTITIES + RELATIONSHIPSBind context and rules.
Use subscription tier, compatible equipment, device model, and setup state to constrain the answer.
GROUNDING + CONSTRAINTSEncode safe resolution.
Name likely causes, respect product limits, and route customers to the correct in-app action or support path.
FAILURE MODES + ROUTINGPrompt authoring · Knowledge architecture · Grounding strategy · Ontology design · Contextual constraints · Failure-mode logic · Response routing
SELF-SERVICE, WITH GUARDRAILS
A useful answer was only half the job.
The experience had to reduce effort without overpromising: explain the issue, expose the limit, recommend the next in-app action, and preserve a path to human help.
Start with approved product truth.
Retrieve only from governed Verizon knowledge and product documentation.
Use the customer's actual setup.
Combine account, equipment, model, and live network context when available.
Translate diagnostics into plain language.
Make the likely issue understandable without burying customers in network terminology.
Recommend a safe next action.
Route to an in-app fix, a documented limit, or human support when self-service should stop.
Context for the shipped assistant—not attributed to my individual contribution.
Observed ticket decrease
Daily VHA help-and-support tickets decreased following release. This is an observed before/after result, not a causal attribution.
Customers with intent signals
Unique customers flagged in production conversations for purchase, upgrade, or product-inquiry intent.
Modeled annual potential
Internal upsell estimate—not booked or realized revenue.
Cost benchmark
Average support-call cost compared with an AI chat; not presented as realized savings.
I connected product strategy, design leadership, prompt architecture, grounded content, safety boundaries, and engineering partnership—without confusing contribution with platform ownership.
Leadership + Organizational Influence
Impact beyond
the interface.
As Verizon Home grew, my role expanded beyond owning critical experiences — strengthening how the organization made decisions and delivered.
CANDOR 2.0 / Analytics as an operating system
Audited, redesigned, handed off and entering phased development.
Platform serving 1M+
That's me!
01 / Set Direction
Translated research, analytics, and technical constraints into scalable IA and roadmap decisions.
554-person Home Security study • 581-person Network MaxDiff
02 / Align the System
Connected functions, leaders, and external partners around one clear customer experience.
Recognized for connecting new departments, leads, and ICs.
03 / Enable the Team
Recruited, onboarded, and mentored two designers, championed VDS, and built a Gen AI learning launchpad.
Both new hires contributing to active projects by week two.
Recognition as Evidence of Trust
Trusted to lead
the hard work.
Recognition followed the way I led — through evidence, alignment, ownership, and an insistence on one clear experience.
Exceptional leadership… championed the need for a single-app experience.
Richard DaltonVP, Experience DesignTitan Award nomination
Innovation + Intellectual Property
Approved
Technology patents
Verizon Titan CX Award
Customer Experience
The trust behind the work
2025 Review"Your strong customer focus, drive for one app… and strong relationships with neighboring departments all played a factor in you winning the Titan Award."
Dave SkuratowiczSenior Director, Home Products & Experiences
and mentoring designers
Closing Reflection
The work shaped the principal designer I became — able to turn ambiguity into direction, evidence into alignment, and craft into organizational leverage.