Katy Simmons

Résumé

Katy SimmonsAI Transformation & Product Executive

Open to AI transformation and product executive roles. Available for select advisory engagements.

Email
katy.simmons@klsconsulting.ai
Phone
(501) 680-9288
LinkedIn
linkedin.com/in/katysimmons
Website
katysimmons.ai
Location
United States

I redesign how companies work with AI. Sixteen years of product leadership taught me how to understand the work before changing it, and I now build the operating models, products, workflows, and AI systems that change how that work gets done.

Download PDF
LinkedIn →Contact →
$20M
ARR, product-led commercial impact
800 → 4,000
customers, proven scale
2 → 436
customers, new-product growth
$1M
first-year revenue, 0→1 business creation
~700K/day
tracks at 99.36% accuracy, AI at production scale
89%
of contracts reconciled automatically end to end
01

Cover letter

How my combined experience in Product and AI shows up in the work

I start with how the work actually happens, identify where the workflow breaks, reimagine the system with the people doing the work, and use evidence to decide what AI should do, what should stay deterministic, and where human judgment still matters.

SoundCloud: Turning an AI model into an end-to-end product experience

SoundCloud needed to identify AI-generated music across roughly 700,000 daily uploads, but detection was only the first Product problem. I owned what a detection actually triggered, including disclosure, review, policy action, monetization and distribution consequences, no action, and appeal.

I designed the evaluation and shipping criteria, tested catalog-representative audio, modeled false-positive cost in artist-trust terms, and set a precision-first tradeoff. I worked closely with Product and Design through research, prototypes, user flows, design reviews, usability testing, and failure and appeal states.

The production system reached 99.36% accuracy, 99.75% precision, roughly 700,000 tracks per day, and 5 to 7 second latency. The work was not deploying a detector. It was turning model output into Product behavior artists and internal teams could trust.

SoundCloud: Reimagining compliance as a product system

SoundCloud faced different age-assurance requirements across the UK, Australia, and other markets. Rather than build separate compliance gates, I worked with Product and Design to create one configurable assurance ladder using the least intrusive method the market and risk allowed, escalating only when more certainty was required.

I translated regulation into Product requirements, modeled friction and drop-off, evaluated methods and vendors, and established a shared decision matrix across Product, Design, Legal, Privacy, and Risk.

The result was one architecture serving multiple regulatory regimes, minimum necessary data collection, and regulatory deadlines met, against UK enforcement exposure of up to £18M or 10% of qualifying worldwide revenue, whichever is greater.

Honeycomb: Reimagining how teams identify AI opportunities

I was hired to identify high-value AI opportunities, but I did not start by asking teams for use cases. I ran 36 discovery sessions, decomposed 100+ workflows into 400+ tasks, designed 23 experiments, and advanced four pilots.

The evidence showed that many supposed AI problems were actually missing-context, workflow-definition, system-connection, ownership, or trust problems. I worked directly with Product and functional teams to replace use-case collection with a repeatable method: observe the work, identify the blocking condition, test narrowly, build what is missing, and grant autonomy only when the evidence supports it.

That became the foundation for Honeycomb's task-assessment framework, autonomy model, GTM operating system, and subsequent Engineering discovery. The evidence did not just validate the strategy. It changed it.

SoundCloud: Turning customer evidence into continuous product discovery

Product could see what changed in telemetry, but the evidence explaining why lived separately across support tickets, bot conversations, Help Center behavior, surveys, and resolution outcomes.

I normalized customer problems across channels, connected bot interactions with downstream ticket outcomes in Looker, and brought qualitative customer evidence into the same BigQuery-centered context layer as product and business data. A PM could move from “adoption fell” to the affected cohort, customer language, recurring problems, failed intents, recent Product changes, and relevant research questions in one investigation.

Combining bot and downstream ticket evidence improved decision accuracy by roughly 20% and contributed to about 2% retention impact.

Honeycomb: Reimagining product discovery around organizational intelligence

Product questions required teams to reconstruct evidence across Gong, Salesforce, Snowflake, product usage, Support, Research, Product and Engineering docs, Marketing, Slack, and external sources.

Rather than build one giant “chat with company data” agent, I designed a persistent intelligence pipeline spanning extraction, normalization, storage, selective retrieval, reasoning, audience-specific output, action, and write-back, while preserving provenance and disagreement. AI handled interpretation and synthesis; SQL, Python, APIs, and automation handled deterministic work.

The result was a shared evidence architecture used across Product, Marketing, Sales, Research, and Support, replacing one-off AI answers with persistent organizational intelligence for recurring decisions.

Additional proof

  • SoundCloud, AI operating modelConsolidated roughly 30 AI pilots across six functions into one roadmap, enterprise governance layer, shared evaluation model, attributable spend, and reusable autonomy patterns.
  • The Weather CompanyReimagined support across three brands serving 100M+ monthly users, reaching roughly 75% automated and self-service resolution, 1M+ estimated annual contacts deflected, 60% fewer duplicate workflows, and 80% less reporting effort.
  • The Weather Company, AI judgmentDefined four interaction classes with explicit confidence thresholds, owners, review rules, and failure paths so automation expanded only when observed performance justified it.
  • SoundCloud, support system redesignRebuilt Help Center, AI, search, routing, escalation, and knowledge as one Product system, reaching roughly 80% automation, 10% lower weekly ticket volume, 90% search success, and fewer than 1% lost tickets.
  • Honeycomb, competitive intelligenceTurned 100+ Gong calls per day into continuous market evidence supporting roughly 40 GTM leaders and early adopters.
02

Experience

  1. May 2026 to Present

    honeycomb.io

    AI Transformation Lead

    • Led 36 discovery sessions across GTM, Finance, Marketing, and Engineering, then decomposed GTM into 100+ workflows and 400+ subtasks so AI decisions happen at task level.
    • Built a task-level AI readiness and autonomy framework (human only, human in the loop, human on the loop, autonomous) now used to decide where AI operates.
    • Personally built the Finance reconciliation system (SQL, Python, JavaScript, Claude skills, MCP, deterministic ARR and iACV logic, evaluation sets, human-review UI, Slack notifications): 89% of contracts reconciled automatically end to end.
    • Launched 4 active AI pilots from 23 identified experiments, each with defined evaluation, ownership, and cost control.

  2. Jan 2025 to May 2026

    The Weather Channel

    Senior Product Manager, AI Transformation (Contract)

    • Unified support across 3 global consumer brands (The Weather Channel, Weather Underground, Storm Radar) into one product model serving 100M+ monthly users, cutting duplicated workflows roughly 60%.
    • Raised automated and self-service resolution to roughly 75%, deflecting an estimated 1M+ support contacts annually while preserving human escalation for high-risk issues.
    • Designed an agent copilot and the AI operating procedures covering source of truth, escalation, and ownership.
    • Built support analytics on containment, CSAT, deflection, and content gaps, routing recurring issues to the product teams that owned them.
    • Reduced support demand roughly 25%, consolidated roughly 70% of common capabilities, cut reporting effort roughly 80%, and reduced contacts from top drivers roughly 30%.

  3. Jan 2024 to Jan 2026

    SoundCloud

    Director of Product, AI Transformation (Contract)

    • Built the company AI operating model across Product, Engineering, Legal, Trust and Safety, Marketing, and Support, including enterprise access, routing, and cost control.
    • Led AI-generated content detection at 99.36% accuracy and 99.75% precision across roughly 700K tracks per day.
    • Reworked support around customer problem resolution at roughly 80K tickets per month, cutting weekly volume 10%; evaluated AI on repeat tickets, escalation, CSAT, content gaps, and bot loops rather than containment, improving decision accuracy roughly 20% with roughly 2% retention impact.
    • Delivered age assurance for UK and Australia regulation, evaluating 3 vendors against a $370K to $860K first-year cost range.

  4. Jan 2019 to Jan 2024

    Apptegy

    Senior Product Manager

    • Conceived and shipped Rooms during COVID, taking it from 2 design partners to 436 customers in 2 years and roughly $5M in sales; it later evolved into Messaging Essentials and Group Connect.
    • Delivered significant 0 to 1 work on Community Experience, Attendance Pro, and Staff Connect.
    • Scaled the flagship SaaS platform 5X, from roughly 800 to 4,000 customers, contributing to roughly $20M ARR.
    • Led 6 global product teams with 6 direct reports, coordinating teams of 4 to 8 across multiple product areas.
    • Established the product strategy and operating models used across the portfolio through scale.

  5. Apr 2015 to Jan 2019

    Arkansas Women's Outreach

    Co-Founder

    • Co-founded a nonprofit delivering period supplies and women's health access in the homeless community.
    • Built the direct distribution model, volunteer operations, and funding partnerships from zero.

  6. Nov 2008 to Dec 2018

    BBA Corp

    Senior Product Manager

    • Launched e-commerce generating $1M in first-year revenue, then scaled it into a multi-campus enterprise product.
    • Established the company's mobile product presence, helping expand the business into the collegiate bookstore market and earn nationwide recognition.
    • Built Product teams and led development of the digital platform as the business expanded.
02

Skills and focus areas

  • AI transformation and strategy: AI transformation, AI strategy, AI readiness assessment, Task-level decomposition
  • AI governance and evaluation: AI governance, Autonomy design, Human in the loop, Model evaluation, AI risk management
  • AI systems and agents: Large language models, LLM deployment, AI agents, Claude, Model Context Protocol, RAG, Prompt design
  • Support automation and knowledge: Support automation, Ticket deflection, Containment rate, CSAT, Self-service
  • Trust, safety, and compliance: Trust and safety, Age assurance, Content moderation, AI-generated content detection, Regulatory compliance
  • People and product leadership: People management, Team leadership, Hiring and coaching, Performance management, Product management, Product operations, Customer discovery, Workflow automation

Tools and platforms: SQL, Snowflake, Salesforce, Looker, Notion, Slack

03

Education

  • 2011 to 2015

    University of Arkansas at Little Rock

    Little Rock, AR

    Bachelor of Computer Science, Computer Systems Engineering

04

Certifications

  • Agentic AI Foundations: Business Applications and Risks

    Harvard Online · Issued July 2026

    Credential ID 260H-EYYC-KWSR

    CertificateVerify →
05

Deeper context

Case studies →Operating principles →Get in touch →