← Selected work

    CitiWatch Cameras / Civic Tech

    A detection is only useful if someone can act.

    Reframing Baltimore’s camera network as a targeted operational data opportunity for illegal-dumping response.

    01 / The tension

    Illegal-dumping reports were largely reactive. Baltimore already had camera infrastructure; the harder question was where better detection could lead to a useful city response.

    I examined camera coverage and geographic incident patterns, and worked with Google AI/ML partners to explore detection feasibility. The work had to connect technical capability with operational priorities, privacy, and the realities of a city service.

    02 / The decisions

    01

    Prioritise places before scaling detection.

    I focused on geographic prioritisation and higher-impact locations, using the existing network as the starting point.

    The trade-off

    Uniform, city-wide automation offered broader coverage, but introduced more technical and governance complexity before the operating model was established.

    02

    Define value at the point of response.

    I framed the recommendations around actionable intelligence for city leadership, rather than model performance alone.

    The trade-off

    A detection capability could demonstrate technical promise without improving an operational decision. The product had to account for the people and processes receiving the signal.

    03 / What I built

    I owned problem framing, camera and incident analysis, geographic prioritisation, opportunity definition, and executive communication. Technical partners and city stakeholders informed the feasibility and recommendations.

    04 / What changed

    Targeted

    coverage framework

    A data-driven approach to prioritising camera coverage and higher-impact geographic areas.

    Explored

    AI-assisted detection

    Analysis and recommendations to inform camera optimisation—not a claim of a deployed city-wide detector or measured incident reduction.

    The lesson I’m taking forward.

    The strongest insight was operational: model capability is one link in the chain. A useful civic product also needs a clear response path and responsible constraints. Technical feasibility matters when it supports an action the city can realistically take.

    A good place to begin

    Something worth
    thinking about?

    I’d love to hear about it. Open to thoughtful conversations about product, systems, and what comes next.