Human Eval

Conduct prompt based evaluations using Google Earth Web and integrated Gemini chat.
Apply detailed rubrics to assess:
Response quality
Geometry correctness (polygons, placements)
Proper naming conventions
Relevance to user queries and map context
Failure modes and edge cases
Complete structured feedback forms with rubric ratings, screenshots, and issue tags.
Log time per evaluation and flag ambiguous or complex queries.
Participate in QA syncs and evaluation calibration sessions.
Maintain consistent throughput while ensuring high-quality outputs.
Tools & Systems
Google Earth, Google Earth Pro (Projects, Polygon, Placemark features)
Gemini Prompt Interface
Internal feedback forms and rater checklists
Google Sheets / Docs for reporting and tracking
Screenshot and logging tools.
Domain & Skill Requirements
Proficient in geospatial analysis: Deep understanding of spatial data, map structures, and geographic context.
Expertise in GIS tools (ArcGIS or Google Earth Pro): Placemarks, polygons, historical imagery, and layer control.
Familiarity with remote sensing data: Experience analysing satellite imagery (historical/current) and vector data formats.
Data annotation & labelling: Ability to assign consistent names, descriptions, and labels to geographic features.
Geospatial visualization: Capable of building spatial maps in Earth Pro to convey insights.
Background in Google Maps, spatial boundaries, and map-based workflows.
Experience in rubric-based assessments and rating scales (e.g. Likert, Pass/Fail)..
Strong analytical and observation skills.
Bonus: Prior experience with LLM evaluation, GIS workflows, map QA, or remote sensing projects