Hologram Routes Fleets Across 550 Networks. Compliance Is the Next Bottleneck.

How Hologram Rewired Cellular Connectivity

A single SIM now reaches 550 carrier networks across 196 countries, according to IoT For All — no contract negotiations, no truck rolls, no border friction. Hologram's Hyper platform, launched in July 2020, turned that spec into infrastructure. Thousands of engineering teams rely on it to keep devices online from Burbank to Brisbane without renegotiating a single carrier agreement.

The old model, including physical SIMs, carrier-by-carrier contracts, and coverage maps that stop at borders, was built for phones in pockets. It was never built for machines that move across continents on their own. As autonomous robots, drones, and sensor fleets deploy into remote and contested environments, the friction of swapping SIMs at every border crossing has become a hard constraint on scale.

Hologram has spent the last decade quietly rewiring that layer. Founded in 2013 as a Chicago startup, the company set out to build what Ben Forgan, its founder and CEO, calls "the operating system for global connectivity." The architecture is deliberately boring in the best way: one SKU, one integration, one bill. Hyper stores multiple carrier profiles on the same eUICC and switches between them remotely as a device crosses coverage boundaries or as Hologram adds new direct carrier relationships. "Software is eating the world," Forgan said at launch. "With IoT, software is also 'eating' the physical world, creating opportunities for companies to build exciting products and generate efficiencies through automation. Hologram is doing our part with Hyper, our eUICC platform, by turning connectivity into software." The result is a connectivity layer that behaves like cloud infrastructure — provisioned programmatically, scaled horizontally, updated without a truck roll.

The problem compounds fast. Within five years of Hyper's launch, billions of devices were projected to communicate over cellular IoT. Refitting each one with a new physical SIM every time it crosses a regulatory boundary or needs a technology refresh is, in Forgan's words, "unwieldy, time-consuming, and economically unfeasible." Hologram's countermeasure: self-improving coverage. As the company expands direct carrier relationships, the profile library on every deployed eUICC grows richer, with no hardware change required. Pricing can be tailored per deployment. Carrier negotiations and integrations happen once, at the platform level, not per fleet.

Early adopters span micromobility fleets, self-driving vehicle programs, medical drone corridors delivering supplies to remote communities, and health monitors operating across six continents. Each deployment generates a stream of connectivity transactions, including handovers, profile switches, and data sessions, that the platform logs, meters, and bills in real time.

That transaction volume is the quiet engine driving the next problem. As fleets scale from hundreds to millions of endpoints, the back office faces a deluge of micro-billing events, tax jurisdictions, and regulatory filings that no spreadsheet can absorb. The connectivity layer has become software. The finance layer has not caught up.

Where the Fleet Operates: Robotics, Defense, Energy, Space

The industry is moving the same direction at the infrastructure layer. On January 28, 2026, Vodafone IoT and Skylo Technologies announced a strategic partnership to integrate satellite-based Non-Terrestrial Network NB-IoT connectivity into Vodafone's global IoT ecosystem, which already supports over 220 million IoT connections worldwide. Skylo's network currently spans 36 countries and covers approximately 70 million square kilometers. The service utilizes the 3GPP Release 17 NB-NTN waveform, allowing it to work with standard IoT hardware and chipsets without requiring specialized satellite dishes or expensive hardware modifications. Customers can manage their entire fleet, whether on cellular or satellite, through the existing Vodafone IoT Managed Connectivity Platform, providing a single interface for global estate management. Erik Brenneis, CEO of Vodafone IoT, said the partnership will "help us deliver ubiquitous connectivity for our customers and make a step-change in network access, reliability, and resilience."

The move addresses a critical bottleneck for industries operating in "dark zones" where terrestrial masts are impractical or non-existent.

That bottleneck is exactly where frontier-tech fleets live. The partnership's own materials call out three verticals that map directly to the autonomous fleet problem set: asset tracking and logistics for high-value goods moving across international borders and through remote territories; agriculture and energy for real-time data from isolated sensors in smart farms, wind farms, and remote utility grids; environmental monitoring for long-term, battery-powered sensors reporting on wildfires, floods, or gas leaks from deep wilderness or oceanic locations. Each of these use cases requires a device that can roam from dense urban LTE to a satellite overlay without human intervention, and each generates the transaction volumes that make per-megabyte carrier billing untenable.

Hologram's Hyper SIM targets the same gap from the device side. Because the SIM carries multiple IMSI profiles and switches profiles over the air, a single hardware build can serve a drone that launches from a U.S. test range, transits international airspace, and lands on a vessel in the South China Sea. The operator never touches a carrier portal. The billing is per-device, not per-carrier, and the data pool is shared across the fleet. That economic model matters when a robotics company ships 500 units to a defense contractor who will deploy them across five continents in the first year — the contractor cannot manage 500 carrier relationships.

The 3GPP Release 17 standard is the enabler nobody talks about. It means the satellite link uses the same NB-IoT modem the device already carries for terrestrial coverage. No separate satellite radio, no patch antenna, no link budget engineering per deployment. The Vodafone-Skylo trial phase is validating performance across diverse geographies now; a full commercial rollout follows. Hologram's platform already abstracts the multi-IMSI switching logic that makes this work on the cellular side. The convergence is structural: the same SIM, the same modem, the same management console, covering terrestrial and non-terrestrial networks.

For a robotics team, this collapses a six-month integration sprint into a configuration change. For a defense program office, it means a single procurement spec covers CONUS and OCONUS deployments. For an energy operator running methane sensors on unmanned platforms in the Gulf, it means the same device that reports over LTE today will report over satellite when the platform moves beyond the shelf — without a firmware update. The infrastructure is no longer the constraint; the application logic is.

When Transactions Bury the Back Office

Every Hyper SIM that Hologram provisions over the air becomes a node in a planetary transaction engine. Multiply thousands of drones, millions of sensors, and nearly 30 billion IoT devices projected for 2030, and the back office confronts a flood of micro-transactions that no spreadsheet can absorb.

Tax complexity is the sharpest edge. Sensors reporting from vineyards, wind turbines, and delivery bots each sit in different VAT regimes, different data-residency rules, and different customs thresholds for cross-border data flows. The European GDPR and California CCPA impose strict requirements on handling sensitive data with full transparency, accountability, and security. Traditional Role-Based Access Control falls short in dynamic environments where access must be context-sensitive. Encryption remains the cornerstone of data security, yet symmetric and asymmetric cryptography are computationally expensive and often unsuitable for real-time applications with large datasets. Homomorphic encryption allows computations on encrypted data without exposing keys, but the computational burden presents a significant barrier in resource-constrained nodes. The IoT's fast growth has revealed the lack of device identity, key management, data integrity, and access control that has existed for a long time in heterogeneous, resource-constrained nodes. Cryptographic overheads for securing endpoints greatly limit the scalability of networks that handle real-time telemetry at scale.

AI and machine learning compound the problem. Models require access to large datasets for training, but anonymization is often insufficient and breaches can occur during processing. Real-time access to data is critical in distributed environments where delays severely impact decisions. The main challenges across the board: scalability, identity management, interoperability, reliability, and security.

Blockchain-based smart contracts can automate financial transactions and reduce processing overhead. Edge computing reduces response time to regulatory violations compared to centrally located cloud architectures. IoT systems improve financial accuracy and simplify compliance management through automated updates to tax laws, audit-ready transaction logs, and regulatory monitoring that tracks compliance metrics to avoid fines.

Yet the research also documents the friction. Blockchain transparency and immutability contradict the principle of data minimization. Traditional Proof-of-Work and Proof-of-Stake are not very effective for constrained IoT scenarios. PBFT-like finality offers better latency at small committees but encounters quadratic messaging overhead with scale. Reputation-supported or lightweight variations lessen chatter but can become less secure under partitions or Byzantine behavior. Energy and green-computing limitations further restrict options, especially for battery-powered nodes.

The compliance crunch is not theoretical. Stripe's own hiring board shows a Business Systems Architect (Tax) role banded at $274,456–$334,600, per Zero G Talent, a signal that fintech infrastructure companies are investing heavily in automated tax logic. The same pressure hits every frontier-tech operator scaling a global fleet.

Automation or Drown: The Back-Office Choice

The economics of global IoT deployment have turned brutal. Insurers' combined ratios are deteriorating — 97.2 percent in 2024, projected to 98.5 percent in 2025 and 99 percent in 2026 — while margins compress across personal and commercial lines. Deloitte's October outlook cites trade policy uncertainty, persistent supply chain disruptions, and labor shortages driving up goods prices and wage inflation. Veteran employees are leaving the workforce faster than recruiting can replace them. The same talent crunch impacts every frontier-tech operator scaling across borders: you cannot hire enough accountants to manually reconcile the transaction volume that a global cellular fleet generates.

Regulators have meanwhile rewritten the rulebook. Over the past two years, more than 170 new regulations have been drafted or passed. Between 2023 and 2024 alone, more than 170 new data protection laws took effect globally. Nearly 40 nations enacted close to 100 data localization policies. The EU's AI Act and Digital Operational Resilience Act (DORA) raised oversight demands further. In the United States, Customs detained an average of 428 shipments per month in 2024 under the Uyghur Forced Labor Prevention Act. Banks paid $19.3 billion in fines in 2024. Financial institutions faced $1.23 billion in penalties in early 2025 alone. The average data breach cost reached $4.88 million in 2024. UK retail and hospitality alone risk £1.6 billion in annual revenue from payment outages. Eighty-six percent of firms increased third-party risk management budgets.

Traditional accounting software was built for a different era. For decades, companies bought tools that helped employees record information: software that digitized the paper ledger but left the judgment to humans. Modern technology companies operate differently: complex subscription models, multiple billing systems, large transaction volumes, international operations, recurring revenue, and increasingly sophisticated financial reporting requirements. The old stack cannot keep pace. Finance teams spend their time moving data between systems instead of analyzing it.

A new generation of platforms aims to close that gap. Relate, an AI-powered accounting and financial automation platform built for modern startups, SaaS companies, and technology businesses, connects directly to the systems where financial activity already happens, including bank accounts, payment processors such as Stripe, billing platforms, and other financial infrastructure. The goal: automate reconciliations, journal entries, invoicing, and revenue recognition. Instead of a monthly close crunch, the concept of continuous close lets automation handle more of the work continuously. AI can connect disparate pieces of information and identify patterns of inconsistency. The long-term vision shifts the paradigm: business activity feeds AI, which produces accounting outcomes for human review. Accounting becomes faster, more continuous, more automated, and potentially much more connected to actual operations.

The market signals confirm the shift. Seventy-seven percent of companies now prioritize AI governance, and surveys show automation reduces compliance costs.

Financial institutions are adopting hybrid data fabric architectures as the new foundation for integration and automation. Retail banks are moving away from flashy generative assistants toward automating unglamorous high-volume processes: onboarding, KYC, compliance checks, internal reporting. Insurance is following suit, with less "AI underwriting" hype and more data accuracy and claims handling automation. Commercial banks are doubling down on decision support and scenario modeling for credit and risk. The FCA's CASS 15 safeguarding rules, coming into force in May 2026, demand a shift from reactive compliance to proactive resilience through continuous reporting and robust data management. Embedded finance in the UK alone is projected to grow from £6.47 billion in 2024 to £15.77 billion by 2029.

The competitive advantage compounds. Automation allows a smaller financial team to manage a larger volume of activity. Every manual process that survives global scaling becomes a headcount tax on the engineering budget. The companies that automate the back office first free capital and talent for the frontier hardware that actually differentiates them.

The barrier is not technical maturity. Gartner predicts over 40 percent of agentic AI projects will be cancelled by 2027. MIT found a 95 percent failure rate for generative AI pilots at enterprises. The main barrier is a misunderstanding of what real transformation requires. Two-thirds of the effort sits in traditional change management: redesigning workflows, redefining roles, helping people adapt. No pilot in a corner delivers that value because it is not designed to transform how the business operates. Real transformation means rethinking roles, responsibilities, and processes from the bottom up. Without that, there are no real P&L gains. The difficult part is not waiting for the technology to mature — it is leading organizational change and ensuring it sticks.

CFOs will demand the same financial discipline from AI initiatives that they expect from any capital investment. 2026 marks the pivot from hype to hygiene: AI's worth proven in operational efficiency, not headlines. Compliance is evolving from a cost center to a competitive advantage. Institutions that can evidence trust will grow fastest.

Engineers Build. Bureaucracy Sustains.

The research is blunt: across two decades of AI in construction, sensor-driven applications, such as safety analytics, progress monitoring, and sustainability modeling, dominate the literature. Administrative and management processes account for 215 papers. That is less than one percent of the total. The gap is not theoretical. "This emphasis has left a major gap: the cognitive and administrative workloads that consume substantial professional time remain largely unaddressed," the Frontiers in Built Environment study concludes.

The same pattern plays out in frontier tech. Hologram's Hyper SIM was built to collapse that overhead: one SIM, provisioned over the air, no carrier contracts, no coverage gaps to plan around. Scale from one device to millions without rebuilding a thing. The connectivity teams at companies no longer file tickets with multiple regional MNOs; they call one API.

The payoff shows up in the field. The Frontiers prototype processed construction specification documents of 800 to 1,500 pages in 6–8 minutes per file on a standard Azure B2ms instance, at under USD 0.10 per file. Practitioners described the outputs as "complete," "consistent," and "a massive value add"; tasks that previously took hours now take minutes. The pipeline is deliberately lightweight: OCR, regular expressions, retrieval-augmented language models. It preserves familiar outputs, such as Excel, Word, and PDF, and keeps humans in the loop where professional judgment is essential. That design philosophy mirrors what Hologram's platform does for connectivity: the engineer sees a single dashboard; the complexity of 550-plus carrier relationships, tax regimes, and regulatory filings stays in the infrastructure layer.

The talent market is signaling the same priority. The Pentagon's "War Force" initiative, announced June 30, 2026, aims to hire hundreds of software engineers and AI specialists to "embed down to the unit level across the department" — not to administer procurement but to accelerate operational capability. Salaries range from $125,776 to $197,200; appointments are term-limited to two years, extendable to four. The application window closed July 24, 2026. GrayMatter's acquisition of New Frontier Technologies, announced June 23, 2026, grew the combined team to more than 500 people across oil and gas, life sciences, food and beverage, distribution, and data-center operations. CEO Paul Galeski framed it as adding "a team of Thinkers and Doers who care about their customers the way we care about ours," a deliberate bet on operational depth over administrative breadth.

On the commercial side, the salary bands tell a parallel story. ASML shows 66 roles added in the past week with a median of $154,000; Stripe shows 80 roles at a $250,000 median. These are hardware, ML, and systems engineers building the physical AI stack.

The longer-term roadmap for administrative AI points toward agentic orchestration: automated routing of submittal packages, reviewer notifications, tighter integration with collaboration tools like Microsoft Teams. For frontier-tech operators, the equivalent is already shipping. Hologram's platform handles the global logistics of connectivity — provisioning, billing, compliance, handover — so the operator's console shows one metric: link health. The divide is not between engineering and administration. It is between work that advances the mission and work that merely sustains the bureaucracy. The companies winning the autonomous fleet race are the ones that push the bureaucracy into the infrastructure layer and keep their engineers on the hardware.

Working in frontier tech? Zero G Talent tracks the openings: see every open ASML role, browse frontier tech jobs, openings at Stripe, and the people building the field.