Mauricio Jimenez

Mauricio Jimenez

Product manager who builds for and with AI

  • Raleigh, North Carolina
  • Authorized to work in the US
  • English & Spanish
Let's talk (LinkedIn, opens in a new tab)

Education

B.S., Industrial Design Engineering
Costa Rica Institute of Technology

Tools

Claude Code, Shopify, Google Analytics, Figma, Optimal Workshop, Notion

How I work

  • Map the whole journey, then add AI

    Before I add AI to a process, I map all of it: every step, every handoff and the people who own them. Speeding up one step rarely changes the result while the steps around it stay as they were. Clarity OS started from the whole planning day, not one feature: before the co-pilot acts, it combines the schedule, goals, conflicts the server has already worked out, and the recent conversation, then acts through 13 typed tools across calendar and tasks. It is an alpha I used myself.

    Proof: Clarity OS, an alpha planning co-pilot, reads the schedule, goals and conflicts together before it acts, through 13 typed tools. See the case: Map the whole journey, then add AI

  • Data-driven, people-centered

    My degree is in industrial design, so I start with people: personas, card sorting and usability tests before a decision hardens. Then evidence keeps the decision honest. For a home-services brand, AI agents audited the tracking and the ad exports and found fake conversions that had steered ad bidding for eight months; the real cost was $946 a lead against a $91 benchmark. With one shared picture of the facts, customers and stakeholders can be weighed together.

    Proof: AI agents found a real cost of $946 a lead against a $91 benchmark, and inflated conversions that had misled ad bidding for eight months. See the case: Data-driven, people-centered

  • Build it to prove it

    I build working prototypes and agents with AI tools, and I set the quality bar before anything ships. To learn which AI models could do real store work, I built an evaluation harness on 18 scenarios taken from real client projects: the answer key is the fix that actually shipped, and objective build and accessibility checks decide pass or fail, not a model's opinion. Localized bug fixes moved to models 10 to 60 times cheaper at reference-equivalent quality, and judgment calls stayed on the strongest model.

    Proof: An evaluation harness of 18 real scenarios, scored by objective checks, showed localized bug fixes could run on models 10 to 60 times cheaper at reference-equivalent quality. See the case: Build it to prove it

Experience

Show:8 of 21 CV items shown: Highlights

  1. Founder

    Jan 2025 – Present

    Divine Power · North Carolina

    • Built a model-evaluation harness on 18 scenarios mined from real client projects, scored against objective build and accessibility gates instead of a model's opinion.
    • Routed production work by task type on those results, moving localized bug fixes to models 10 to 60 times cheaper at reference-equivalent quality.
    • Exposed a real $946 cost per lead (vs a $91 benchmark) and a projected ~$700K opportunity by having AI agents audit a brand's tracking, uncovering inflated conversions that had misled ad bidding for 8 months.
    • Built a multi-agent operation orchestrated with Paperclip over a shared knowledge base, keeping agent output consistent across the business.
    • Built tool-using agents that act on live APIs, including a text-to-SQL finance agent and a Shopify agent that validates against the schema before any write.
    • Designed and built Clarity OS with AI-assisted development: a planning assistant whose agent acts on calendar and tasks through 13 tools (alpha).
    • Specified a Shopify product configurator that allows only valid builds based on real-time inventory, and built the store's product-description tool with Claude.
    • Built an AI content pipeline for Shopify that extracts product data, rewrites copy on-brand and batch-tests quality before re-uploading.
    • Surfaced a ~50% conversion opportunity for a DTC brand, a path to ~2x revenue from existing traffic, by auditing the full funnel and reprioritizing the roadmap.
    • Launched a DTC brand on a tight buyer persona and a clear pivot-or-reinforce decision framework, hitting every month-one target.
    • Wrote market and positioning strategies for brands doing six to eight figures in revenue, from market sizing to quick-win actions.
    • Ran competitive analysis and formal usability testing through card sorting and heatmaps.

  2. Head of Product & Growth

    Oct 2022 – Aug 2024

    Specialized Bicycle Components · Costa Rica

    • Led the data overhaul behind the business's BI across 15,000+ customers and 10,000+ SKUs, cutting reporting lag from two-to-three months to near real-time.
    • Won the budget and cleared the roadblocks for that build, moving the BI team off manual cleanup and giving leadership its first accurate view of product, inventory and sales.
    • Grew the e-bike category ~4x year over year during an industry-wide downturn by owning pricing, product mix, distribution and retail-channel expansion.
    • Ran a price-elasticity analysis on top sellers and renegotiated supplier cost prices against projected sales and margin targets.
    • Led teams of four to five people across product and marketing.
    • Presented performance, plans and business cases directly to regional and global CFOs.
    • Negotiated agreements that brought multiple international brands into the regional distribution portfolio.

  3. Associate Product Manager

    Sep 2021 – Sep 2022

    Specialized Bicycle Components · Costa Rica

    • Launched BikeStation's ecommerce from scratch on Shopify in three months, leading the design and the ERP integration.
    • Owned the Volcano 100 registration platform from requirements to launch, leading delivery to 13% checkout conversion and record sales; it still runs every year.