← All experience

At a glance

Your history with Harborlight Software

B2B software company that helps local service organizations coordinate cases, appointments, and customer communication.

Saved evidence

Everything below remains connected to this organization.

  • Roles and projects 2
  • Work updates 2
  • Open recommendations 0

Roles and projects

What you did

Each record keeps its responsibilities, achievements, skills, tools, methods, and open questions together.

Role

Lead Product Manager

Direct

–Present · Portland, Oregon · Hybrid · Full Time

Leads discovery and delivery for case-management and communications workflows used by local-service teams.

Responsibilities

  • Own product strategy, quarterly outcomes, discovery, and roadmap decisions for two cross-functional squads.
  • Partner with customer operations, design, engineering, privacy, and support on workflow improvements.
  • Define responsible AI evaluation criteria and human-review safeguards for an assisted case-summary pilot.

Achievements

  • Reduced median case-intake time by 31% after redesigning forms and routing rules.
  • Improved successful self-service completion from 62% to 78% through research-led onboarding changes.
  • Launched an AI-assisted summary pilot with opt-out controls, audit logs, and weekly quality review.

Skills

  • Product strategy
  • Product discovery
  • Roadmapping
  • Responsible AI
  • Service design
  • Product analytics
  • Stakeholder management

Tools and technologies

  • Figma
  • SQL
  • Amplitude
  • Jira
  • REST APIs

Methods

  • Customer interviews
  • Opportunity solution trees
  • A/B testing
  • Outcome-based roadmaps

Industry experience

  • SaaS
  • Civic technology

Project

AI Intake Pilot Product Lead

Direct

· Remote · Remote · Internal Project

Led a bounded pilot that suggested case classifications while keeping final decisions with trained staff.

Responsibilities

  • Define pilot scope and success measures.
  • Coordinate model evaluation, privacy review, and frontline training.
  • Run weekly error reviews and prioritize remediation.

Achievements

  • Reached 84% reviewer agreement on eligible cases in the evaluation set.
  • Cut manual triage effort by 18% without automating final eligibility decisions.

Skills

  • AI product management
  • Experiment design
  • Model evaluation
  • Change management
  • Privacy by design

Tools and technologies

  • Python notebooks
  • OpenAI API
  • PostgreSQL

Methods

  • Human-in-the-loop review
  • Pilot design
  • Error taxonomy

Industry experience

  • Civic technology
  • Artificial intelligence

Recent evidence

Work updates

Specific wins, projects, feedback, problems solved, and milestones appear here with all saved supporting detail.

Achievement

Added weekly AI quality review

Direct

Created a recurring review with frontline staff to inspect assisted summaries, categorize errors, and decide whether the pilot should expand.

Scope

  • AI pilot governance

Skills

  • Responsible AI
  • Model evaluation
  • Facilitation

Outcomes

  • The pilot documented three recurring error patterns and resolved the highest-impact issue before wider rollout.

Work Update

Reframed roadmap around service outcomes

Direct

Replaced a feature list with measurable outcomes for intake time, self-service completion, and staff rework.

Scope

  • Product operating model

Skills

  • Roadmapping
  • Product strategy
  • Stakeholder management

Outcomes

  • Quarterly planning discussions became shorter and tradeoffs were easier to explain.

Details to strengthen

Useful follow-up questions

These recommendations come from missing details in the roles and updates above. They are prompts for clarification, not evidence by themselves.

No open recommendations are saved for this organization.