Proximity AI

Cover Block

PUBLIC

Field	Value
Name	Proximity AI
Tagline	Developing ontology and world models to connect physical data sources for real-time digital twins of planetary activity.
Industry	Deeptech
Technology	AI / Machine Learning
Geography	North America

Links

PUBLIC

• Website: https://www.useproximity.ai/
• LinkedIn: https://www.linkedin.com/company/proximityai

Executive Summary

PUBLIC

Proximity AI is an early-stage deeptech company building what it describes as the ontology and world models needed to connect thousands of physical data sources simultaneously, with the stated goal of producing a real-time digital twin of planetary activity [useproximity.ai]. The pitch sits at the intersection of three investor-relevant currents: the emergence of physical-world AI agents, the proliferation of sensor and satellite data, and the growing appetite for ground-truth data layers that machine learning systems can actually act on. According to the company's public site, the platform is intended to serve humans, AI agents, and robots as end consumers of that data, with a reference to a satellite component labelled SAT-01 [useproximity.ai]. A third-party listing characterizes the platform as one that connects diverse streams including permits, sensors, supply chains, public data, cameras, and satellites [AgentsPointee]. Beyond these descriptions, the public footprint is thin: founders, headquarters, founding year, funding history, and team composition are not disclosed in the sources captured for this report. The most useful public signal over the next 12 to 18 months will be whether Proximity AI moves from a stated mission into named pilot customers, a disclosed funding round, or a published technical artifact (an SDK, a dataset, an API) that lets outside developers verify the ontology claim. Until then, the company should be treated as a thesis worth tracking rather than a position worth sizing.

Data Accuracy: YELLOW -- Confirmed by company website and one third-party listing; founder, funding, and headcount data not publicly available.

Taxonomy Snapshot

Axis	Value
Industry / Vertical	Deeptech
Technology Type	AI / Machine Learning, ontology and world models
Geography	North America

Company Overview

PUBLIC

Proximity AI presents itself publicly through a single-page site under the tagline "Scrape The Reality," with the stated mission of "developing the ontology and world models to connect thousands of physical data sources simultaneously, creating a real-time digital twin of planetary activity" to serve humans, agents, and robots [useproximity.ai]. The site references a satellite or sensor node labelled SAT-01 with a coordinate reading at latitude 34.0522 N and longitude 118.2437 W, which corresponds to Los Angeles [useproximity.ai]. Whether that coordinate reflects a headquarters, a demo node, or a stylistic flourish is not clarified in any captured source.

The company maintains a LinkedIn company page at /company/proximityai, but the page does not surface a public description, employee count, or location in the captured snippets [LinkedIn]. A separate page at /company/getproximity exists under a similar handle and may or may not be related; the captured material does not disambiguate the two [LinkedIn]. Founding date, legal entity, incorporation state, and founding team are not publicly available in the sources reviewed for this report. There is also a separately operated firm called Proximity Works at proximity.tech, led by Hardik Jagda, which is a distinct entity and should not be conflated with Proximity AI [proximity.tech].

Given the absence of disclosed milestones (no announced funding, no named customers, no press coverage in major outlets captured here), the most accurate characterization is that Proximity AI is operating in a stealth or pre-announcement posture. Investors evaluating the company at this stage should expect to source primary information directly from the founders rather than from public databases.

Data Accuracy: ORANGE -- Single-source company website plus unpopulated LinkedIn page; no third-party press confirmation of company history.

Product and Technology

MIXED

The product, as described publicly, is a data-integration and world-modeling layer rather than an end-user application. The company's own framing centers on "ontology and world models" capable of connecting thousands of physical data sources simultaneously to produce a real-time digital twin of planetary activity [PUBLIC] [useproximity.ai]. A third-party catalog entry expands the description, listing the input data streams as "permits, sensors, supply chains, public data, cameras, satellites, and more," and frames the output as real-time awareness of the physical world [PUBLIC] [AgentsPointee]. Taken together, the public material points to a platform whose value proposition rests on breadth of ingestion (many heterogeneous physical-world feeds) and on a unifying semantic layer (an ontology) that allows downstream agents to reason across those feeds.

Two technical claims warrant flagging. First, the reference to SAT-01 implies either an owned, leased, or partnered satellite asset, but no specifications, launch partner, or operator are disclosed in any captured source [PUBLIC] [useproximity.ai]. Second, the phrase "world models" is doing meaningful work in the pitch: in current AI research, world models typically refer to learned simulators of an environment that an agent can plan against. Whether Proximity AI is training such models in-house, fine-tuning open base models, or applying the term more loosely to a knowledge graph is not specified in the public material.

No SDK, API documentation, pricing page, customer logo wall, or developer portal is surfaced in the captured sources. There is a Zapier integration listing referencing "Proximity AI by Zapier," but the listing in the captured snippet describes Zapier's generic AI integration capabilities rather than confirming a productized Proximity AI connector with disclosed actions or triggers [PUBLIC] [Zapier]. Tech-stack details, hosting choice, and model architecture are not publicly available.

Data Accuracy: ORANGE -- Product description rests on company self-description and one third-party listing; no independent technical validation surfaced.

Market Research and Opportunity

PUBLIC

The market Proximity AI is targeting matters now because two adjacent waves are converging: AI agents that need grounded, real-time inputs to act in the physical world, and a sensor and satellite economy that is producing more raw observation data than existing platforms can semantically reconcile.

The captured research package does not include named third-party reports sizing the digital-twin, geospatial-intelligence, or world-models markets, so any TAM figure in this section would be an unsourced estimate and is therefore omitted. What can be said from the public framing alone is that Proximity AI's stated input set (permits, sensors, supply chains, public data, cameras, satellites) spans at least four distinct buyer categories that already have established procurement budgets: geospatial analytics buyers (defense, intelligence, insurance, agriculture), supply-chain visibility buyers (logistics, retail, manufacturing), smart-infrastructure buyers (utilities, municipalities, real estate), and the emerging robotics and autonomous-systems segment that needs world models for planning [AgentsPointee]. Each of these segments has its own incumbents and procurement cycles, which is both an opportunity (multiple wedge entries) and a risk (no single beachhead).

Demand drivers visible in the public framing include the maturation of foundation models that can consume multi-modal inputs, the falling cost of satellite imagery and IoT sensors, and the explicit positioning of the platform to serve "agents and robots" as buyers, not just humans [useproximity.ai]. The agent-as-customer framing is notable because it implies a machine-to-machine API revenue model rather than a seat-based SaaS model, which historically scales differently and benefits from usage-based pricing.

Regulatory and macro forces cut both ways. Geospatial and sensor-fusion platforms increasingly operate in jurisdictions with data-sovereignty requirements, satellite-imagery export controls in the United States, and growing scrutiny of AI systems that ingest public surveillance data. The same forces also create demand for compliant, auditable data layers, which a well-architected ontology platform can serve.

Cited market signal	Source
Platform connects "thousands of physical data sources simultaneously"	[useproximity.ai]
Input streams include permits, sensors, supply chains, public data, cameras, satellites	[AgentsPointee]
End consumers framed as humans, agents, and robots	[useproximity.ai]

The analyst takeaway is that Proximity AI is positioning into a market that is real, large, and currently fragmented across at least four buyer segments, but the captured public material does not yet anchor the opportunity to a specific wedge or a named third-party sizing report.

Data Accuracy: YELLOW -- Market framing inferred from company and third-party descriptions; no named analyst report on TAM captured.

Competitive Landscape

MIXED

Proximity AI is positioning into a category where established geospatial-intelligence vendors, digital-twin platforms, and supply-chain visibility specialists already operate, and the captured sources do not name a direct head-to-head competitor [PUBLIC].

Because no competitor is named in the structured facts, the competitive analysis here is necessarily framed at the segment level rather than the company level. The segments most relevant to Proximity AI's stated scope fall into three buckets. The first is geospatial and Earth-observation analytics, where publicly known players (not cited in this research package and therefore not enumerated as facts here) have spent a decade building satellite tasking, imagery pipelines, and analytic overlays for defense and commercial buyers. The second is the digital-twin and industrial-IoT segment, historically dominated by large industrial-software vendors that sell facility-scale or city-scale twins rather than planetary-scale ones. The third is the emerging "world models for agents" research frontier being pursued by frontier AI labs and robotics-focused startups, which overlaps with Proximity AI's stated technical ambition.

Where Proximity AI could establish a defensible edge, based on its public framing, is at the semantic layer: an ontology that normalizes heterogeneous physical-world feeds is harder to replicate than any single data pipeline, and it benefits from compounding (every new source type makes the ontology more useful for the next) [PUBLIC] [useproximity.ai]. That edge is durable only if the ontology is genuinely proprietary and if customers integrate it deeply enough to make switching costly. It is perishable if open-source ontologies, foundation-model providers, or large cloud platforms ship a comparable abstraction as a free layer beneath their own services.

Where the company is most exposed is in distribution and capital. Geospatial incumbents have multi-year government contracts and embedded sales motions. Industrial-twin incumbents have decades of relationships with operators of factories, ports, and grids. Without a disclosed funding round or named anchor customer, Proximity AI's path into those buyer relationships is not yet visible in public material. The most plausible 18-month scenario is bifurcated: the company wins if it lands a marquee design partner in either defense, supply-chain, or robotics that validates the ontology in production; it stalls if it remains a horizontal platform without a named wedge while better-capitalized vertical specialists ship narrower but shippable products.

Data Accuracy: ORANGE -- No named competitors in captured sources; segment framing is analyst inference.

Opportunity

PUBLIC

If Proximity AI executes on its stated mission, the prize is to become the default semantic layer between the physical world and the AI systems that act on it.

The headline opportunity. The single largest outcome Proximity AI could plausibly become is the connective tissue that lets AI agents and robots reason about the physical world in real time, the way mapping APIs became the connective tissue for location-aware software in the prior decade. The company's own framing ("ontology and world models to connect thousands of physical data sources simultaneously, creating a real-time digital twin of planetary activity") is consistent with that ambition [useproximity.ai]. The third-party characterization of the platform as one that ingests permits, sensors, supply chains, public data, cameras, and satellites suggests the team is not picking a single vertical but building horizontal infrastructure [AgentsPointee]. Horizontal infrastructure is harder to sell early but, when it works, it captures a tax on every vertical built on top.

Growth scenarios.

Scenario	What happens	Catalyst	Why it's plausible
Agent-infrastructure wedge	Proximity AI becomes the grounded-data API that AI agent frameworks call when they need real-world state	A frontier-model lab or major agent framework integrates the platform as a default tool	Company explicitly positions "agents and robots" as end users [useproximity.ai]
Defense and dual-use anchor	Platform lands a defense or intelligence design partner that validates planetary-scale fusion	A pilot contract with a defense innovation unit or allied government	Stated input set already includes satellites and public data, the core defense fusion stack [AgentsPointee]
Supply-chain operating layer	Platform becomes the live-state layer for global logistics and trade	A top-tier shipper, retailer, or 3PL signs a multi-year data agreement	Supply chains are explicitly named as an ingest source [AgentsPointee]

What compounding looks like. The flywheel for an ontology business is straightforward in theory: each new connected data source adds value not only for its own buyers but for every buyer of every other connected source, because the joins across sources are where unique insight lives. If Proximity AI can sign three or four anchor data partners across satellite, sensor, and permit categories, the marginal cost of adding the fifth and sixth drops while the marginal value rises. The captured public material does not yet show evidence that this flywheel has started, so the compounding case remains a thesis rather than a measured trend.

The size of the win. Public comparables in adjacent categories (geospatial analytics platforms, industrial digital-twin vendors, mapping infrastructure providers) have historically reached multi-billion-dollar enterprise values when they captured a defensible position in their layer of the stack. The captured sources do not include a named TAM report or a specific public comparable for the planetary-twin category, so any specific dollar figure here would be unsourced. Stated qualitatively (scenario, not a forecast): if Proximity AI becomes the grounded-data API for AI agents, the outcome resembles infrastructure-layer comparables rather than vertical SaaS comparables, and the company would be valued accordingly.

Data Accuracy: YELLOW -- Opportunity framing rests on company self-description plus analyst-constructed scenarios; no revenue, customer, or funding evidence captured to anchor magnitude.

Sources

PUBLIC

1. [useproximity.ai] Proximity AI - Scrape The Reality | https://www.useproximity.ai/

2. [AgentsPointee] Proximity AI: Reviews, Alternatives, Pricing, & Use Cases | https://agentspointee.com/listing/proximity-ai/

3. [Shapes AI] proximity ai | Shapes AI | https://shapes.inc/proximityai

4. [LinkedIn] Proximity AI company page | https://www.linkedin.com/company/proximityai

5. [LinkedIn] Proximity (getproximity) company page | https://www.linkedin.com/company/getproximity

6. [Zapier] Proximity AI by Zapier Integration - Quick Connect | https://zapier.com/apps/proximity/integrations/ai

7. [proximity.tech] Proximity Works - AI Systems Built for Scale (separate entity, included for disambiguation) | https://www.proximity.tech/