Associate Geospatial Analyst

Who We Are

4M Analytics is building the first Subsurface World Model: an AI platform that uses visual reasoning and large-scale data fusion to turn the undocumented world beneath our streets into one living, intelligent map. We make reliable utility data accessible to the construction industry, government agencies, and utility owners, and we are changing how infrastructure is understood and protected.

We are Series C, backed by Insight Partners, Viola, and F2, with 120 people across Tel Aviv and Austin, Texas. The problem is massive, the momentum is real, and we are moving fast.

The Opportunity

Mapping a utility network starts with everything we know about what is under a street: utility records, as-built plans, imagery, detected objects and field evidence. Our AI line generation takes all of that and produces the network. But a model alone does not decide what ships. There is a human dialog with it: a geospatial expert who reviews what it produced, refines it, fills in what is missing, corrects what is wrong, and decides when it is right enough for a customer to rely on.

That dialog is the job, and it is getting deeper. Today the expert does more of the correcting by hand. As the AI takes over more of the generation, the expert's time moves to judging the output, measuring where the model fails, writing the mapping rules and training examples it learns from, and acting as the utility domain expert our data and AI teams build with. The people who know how utilities actually run under a street are the ones who will teach the model to map them.

You will start hands-on mapping and refining utility networks (power, gas, water, telecom and more) for infrastructure projects across the US, working with our Tel Aviv team and our offshore mapping team. As automation covers more ground, your time shifts to QA, agent training and domain expertise. The end state for this role is a utility domain expert who helps build the tools, corrects the models, and holds the quality bar for what the AI produces.

What You’ll Do

Refine what the AI produces, then judge and teach the model

• Review AI-generated utility networks against records, plans and imagery; refine, complete and correct them, and decide what is ready to ship

• Digitize and enrich utility lines and features in QGIS, and query, clean and validate spatial data in PostGIS

• Run QA on mapped data for completeness, accuracy and sector coverage before it goes to customers

• Compare estimated vs. actual mapping results per project and report the gaps

• Review the offshore mapping team's output and give clear, specific feedback

• Document how you map as SOPs the team can follow

• QA auto-generated lines against manual mapping and flag errors, gaps and wrong sectors

• Track accuracy and coverage per area and use the numbers to show the data and AI teams where automation needs work

• Write the mapping rules, examples and training data our AI agents learn from, and give structured feedback on agent output so it improves week over week

• Test and roll out model and tooling improvements together with the data and AI teams

• Be the utility domain expert in the room when the tech team decides what to build next

What You Bring

• Motivation and Excellence. You've earned a bachelor's degree (any field), OR have hands-on GIS experience.

• An analytical eye. You catch the wrong line, the missing manhole and the mislabeled sector, whether a person or a model drew it, and you can explain why it is wrong.

• Data-driven. You measure accuracy and coverage, track quality metrics, and use the numbers to say where the work needs to go, not just that something looks off.

• Clear documentation. You can turn the way you map into rules and examples that another analyst, or an AI agent, can follow.

• Independent. You run several projects at once, manage your own queue, and ask for help early when a project is at risk.

• English. Good written and spoken English. You will work on US projects and with an offshore team.

AI-First Mindset

We are building an AI-first company. We expect every analyst, along with the rest of the company, to use AI tooling daily for mapping, QA, analysis and documentation.

You do not need to build models. You do need to want to work next to them, be curious about why they fail, and be willing to become the person who teaches them.

Standout Attributes

• Familiarity with GIS tools, architecture, civil engineering, or similar

• Familiarity with utilities, telecom or infrastructure

• Familiarity with computer vision on imagery (object detection, for example)

• Basic skills in Python, SQL, or BI