Bottleneck Map

Where founders, capital, and public demand fail to meet.

Atomic answer

Sovereign technology does not fail because actors are absent. It fails because founders cannot see which government buyers need their technology, capital allocators cannot evaluate technical claims at the depth their mandates require, strategic buyers cannot find qualified suppliers, and regulators cannot read the product roadmaps their rules are reshaping. This map tracks 25 bottlenecks where these failures concentrate - across 10 layers, with signal strength and source confidence for each.

What the map tracks

This is an analytical map, not a database. Each bottleneck is a pattern Stack & State has observed where actors in the sovereign technology ecosystem cannot see each other's constraints in time. Bottlenecks are classified by signal strength (how visible and urgent the pattern is) and source confidence (whether the assessment draws on primary institutional sources, mixed public and analytical evidence, or Stack & State ecosystem analysis).

Layer 1: Materials & Processing

Layer 1: Materials & Processing

HIGH MIXED

Rare earth separation capacity without bankable offtake

Controls

China processing dominance

Needs

Allied separation facilities with long-term purchase commitments

Why now DPA Title III can start capacity; offtake determines survival

Source note Mixed public-source and analytical assessment. Treat as source-reviewed directionally, not as a quantified finding.

Last reviewed 2026-06-12

What would change this assessment Visible long-term purchase agreements, project-finance commitments, or public buyer commitments would reduce the bottleneck.

Layer 1: Materials & Processing

HIGH PRIMARY

Graphite spheronization concentrated in single country

Controls

China-dominant synthetic and natural graphite processing

Needs

Battery supply chain diversity

Why now Every EV and grid battery depends on it

Source note IEA critical minerals reporting and public graphite supply-chain analysis.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Layer 2: Energy & Power

Layer 2: Energy & Power

HIGH MIXED

AI data center power demand outpacing utility planning cycles

Controls

Hyperscaler procurement

Needs

SMR licensing reform, transmission buildout

Why now Data center load growth is reordering utility capital plans

Source note Mixed public-source and analytical assessment. Treat as source-reviewed directionally, not as a quantified finding.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Layer 2: Energy & Power

MEDIUM PRIMARY

Grid interconnection queues blocking energy projects

Controls

FERC/RTO interconnection processes

Needs

Queue reform, faster study cycles

Why now Projects can face multi-year interconnection delays

Source note FERC interconnection reform materials and public interconnection queue reporting.

Last reviewed 2026-06-12

What would change this assessment Queue reform, faster study completion, or material reductions in interconnection backlogs would reduce the bottleneck.

Layer 3: Compute & AI Infrastructure

Layer 3: Compute & AI Infrastructure

HIGH ANALYTICAL

Founders cannot map FedRAMP/IL5 requirements to architecture decisions

Controls

DoD cloud security requirements

Needs

Builder-readable compliance translation

Why now Cloud choice determines market access before a line of code ships

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 3: Compute & AI Infrastructure

MEDIUM MIXED

Sovereign AI mandates creating fragmented compliance surface

Controls

EU AI Act, US executive orders, national AI strategies

Needs

Unified compliance frameworks

Why now Companies selling to multiple governments face incompatible requirements

Source note Mixed public-source and analytical assessment. Treat as source-reviewed directionally, not as a quantified finding.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Layer 4: Autonomy & Robotics

Layer 4: Autonomy & Robotics

HIGH MIXED

Export-control sensitivity blocking allied procurement of autonomous systems

Controls

ITAR/EAR classification

Needs

Allied-accessible product architectures

Why now Systems are often over-classified relative to capability

Source note Mixed public-source and analytical assessment. Treat as source-reviewed directionally, not as a quantified finding.

Last reviewed 2026-06-12

What would change this assessment Evidence of repeatable transition paths from prototype awards to procurement would change the assessment.

Layer 4: Autonomy & Robotics

MEDIUM ANALYTICAL

Defense primes absorbing innovation without scaling it

Controls

Prime contractor acquisition processes

Needs

Faster SBIR-to-program-of-record pathways

Why now Startups prototype faster than primes can integrate

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 5: Export Controls & Compliance

Layer 5: Export Controls & Compliance

HIGH PRIMARY

Deemed export rules trapping foreign-national technical talent

Controls

ITAR §120.17, EAR §734.13(b)

Needs

Clearer rules for dual-use software

Why now Defense startups struggle to hire globally

Source note Public export-control rules: ITAR section 120.17 and EAR section 734.13(b).

Last reviewed 2026-06-12

What would change this assessment Clearer agency guidance, licensing pathways, or product-architecture standards would change the assessment.

Layer 5: Export Controls & Compliance

HIGH MIXED

AI model-weight controls undefined for most founders

Controls

BIS advanced computing rules

Needs

Clear classification thresholds for AI

Why now Founders are building in regulatory uncertainty

Source note Public BIS advanced-computing rulemaking and AI governance materials.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Layer 5: Export Controls & Compliance

MEDIUM ANALYTICAL

Entity-list screening not automated at startup scale

Controls

BIS Consolidated Screening List

Needs

Compliance-as-code tools

Why now Manual screening fails after 10 customers

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 6: Non-Dilutive Capital

Layer 6: Non-Dilutive Capital

HIGH MIXED

SBIR Phase II awards without transition sponsors

Controls

DoD program offices

Needs

Program-office champions before Phase II ends

Why now Phase III transition remains a recurring gap for companies without a committed government customer

Source note SBA, SBIR.gov, DoD SBIR/STTR materials, and GAO reporting on transition challenges.

Last reviewed 2026-06-12

What would change this assessment Evidence of repeatable transition paths from prototype awards to procurement would change the assessment.

Layer 6: Non-Dilutive Capital

HIGH MIXED

DPA Title III funding capacity without offtake commitments

Controls

DoD industrial base policy

Needs

Strategic buyer purchase agreements

Why now Public capital can start capacity; only offtake makes it permanent

Source note Public DPA Title III program materials and industrial-base policy releases.

Last reviewed 2026-06-12

What would change this assessment Visible long-term purchase agreements, project-finance commitments, or public buyer commitments would reduce the bottleneck.

Layer 6: Non-Dilutive Capital

MEDIUM ANALYTICAL

OSC loan authority under-accessed by hardware startups

Controls

Office of Strategic Capital

Needs

Simpler application pathways

Why now Hardware-heavy startups need patient capital between equity and project finance

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 7: Government Buyers

Layer 7: Government Buyers

HIGH MIXED

Program offices evaluating AI without technical evaluation capacity

Controls

DoD acquisition workforce

Needs

Market-side signal and technical co-evaluation

Why now Large AI and autonomy procurements can be difficult for program offices to evaluate independently

Source note Mixed public-source and analytical assessment. Treat as source-reviewed directionally, not as a quantified finding.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Layer 7: Government Buyers

MEDIUM ANALYTICAL

Innovation units duplicating company discovery across silos

Controls

DIU, AFWERX, NSIN, DARPA, service labs

Needs

Shared qualified-company pipelines

Why now Founders pitch the same technology to five different offices

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 8: Strategic Capital & CVCs

Layer 8: Strategic Capital & CVCs

HIGH ANALYTICAL

CVCs unable to diligence dual-use technology at required depth

Controls

Corporate venture investment committees

Needs

Technical co-evaluation and policy-risk assessment

Why now CVCs can bridge public demand signals and private scaling capital when diligence covers both technology and regulatory surface

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment Public examples of repeatable technical diligence processes or dedicated sector teams would change the assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 8: Strategic Capital & CVCs

MEDIUM ANALYTICAL

Corporate buyers without visibility into startup supply chains

Controls

Prime contractor procurement

Needs

Pre-vetted supplier pipelines

Why now Primes need qualified subcontractors but cannot evaluate hundreds of startups

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 9: Sovereign & Allied Capital

Layer 9: Sovereign & Allied Capital

HIGH ANALYTICAL

Sovereign wealth funds deploying into strategic tech without technical diligence

Controls

SWF investment mandates

Needs

Layer-specific technical evaluation frameworks

Why now Strategic technology allocations often require technical diligence outside a generalist investment process

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment Public examples of repeatable technical diligence processes or dedicated sector teams would change the assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 9: Sovereign & Allied Capital

MEDIUM ANALYTICAL

Development finance institutions lacking project pipeline visibility

Controls

DFC, IFC, AfDB lending criteria

Needs

Bankable project pipelines

Why now DFIs have mandates but cannot find projects at scale

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 9: Sovereign & Allied Capital

MEDIUM MIXED

Allied industrial policy creating incompatible supplier requirements

Controls

AUKUS Pillar II, EU Critical Raw Materials Act

Needs

Interoperable supplier qualification standards

Why now Companies selling to multiple allies face conflicting rules

Source note Public AUKUS Pillar II and EU Critical Raw Materials Act materials.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Layer 10: Founders & Industrial Operators

Layer 10: Founders & Industrial Operators

HIGH ANALYTICAL

Technical founders building into regulatory frameworks they cannot read

Controls

ITAR, EAR, CFIUS, EU AI Act

Needs

Builder-readable regulatory translation

Why now Late regulatory discovery can force expensive product and market-access changes

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 10: Founders & Industrial Operators

HIGH MIXED

Startups pursuing SBIR/STTR without transition planning

Controls

Agency SBIR programs

Needs

Transition memoranda before prototype start

Why now Programs with transition planning have a clearer path from prototype funding to procurement

Source note SBA, SBIR.gov, DoD SBIR/STTR materials, and GAO reporting on transition challenges.

Last reviewed 2026-06-12

What would change this assessment Evidence of repeatable transition paths from prototype awards to procurement would change the assessment.

Layer 10: Founders & Industrial Operators

MEDIUM ANALYTICAL

Dual-use founders unable to price the security premium

Controls

Government procurement pricing rules

Needs

Market data on security-premium pricing

Why now Founders under-price government contracts or avoid them entirely

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Layer 10: Founders & Industrial Operators

MEDIUM ANALYTICAL

Labs with commercializable IP lacking commercialization pathways

Controls

Federal lab tech transfer offices

Needs

Founder, strategic buyer, or program office connections

Why now Federal labs produce substantial technical work that often needs a clearer commercialization route

Source note Stack & State analytical judgment based on ecosystem observation. No public source is asserted for this classification.

Last reviewed 2026-06-12

What would change this assessment New public evidence, corrections, or operator signals that materially change the constraint would update this assessment.

Analytical judgment: this classification is a navigation hypothesis, not a verified factual finding.

Structural gaps

Five patterns emerge across the map.

Translation

Founders cannot read regulation. Buyers cannot read technology. The gap is the market.

Visibility

Qualified companies cannot find buyers. Buyers cannot find qualified companies. The map is the product.

Diligence

Capital moves faster than the ability to evaluate what it is buying. Technical co-evaluation is the scarcest service.

Offtake

Public capital starts capacity. Only purchase commitments make it permanent. The offtake gap is the bridge between policy and production.

Classification

Products are over-classified or under-classified relative to capability. Export-control architecture decisions shape market access before a line of code ships.

How to use this map

Each audience reads a different layer.

For founders

Trace your product's bill of materials across the layers. Where does your supply chain touch a bottleneck? That is where your regulatory surface and your procurement opportunity live.

For CVCs

Map your deal flow against the bottleneck layers. The highest-value investments sit where public demand, technical capability, and capital scarcity converge.

For government buyers

Identify the bottlenecks in your acquisition layer. The qualified companies you cannot find are the ones founders cannot route to you.

For sovereign/allied capital

Use the signal-strength indicators to prioritize where technical diligence will produce the highest-conviction investment decisions.

Related reading

Articles and tools for each layer.

Layer 1: Materials

The Materials Bottleneck Behind Sovereign Technology

Processing, separation, offtake, and permitting -- the real constraint in critical minerals.

Layer 5: Export Controls

Export Controls for Deep-Tech Startups

ITAR, EAR, CFIUS, and entity-list constraints translated into product architecture decisions.

Layer 6: Non-Dilutive Capital

Non-Dilutive Capital for Defense Tech

Every program, decision tree, and transition strategy for founders navigating government funding.

Layer 6: Capital

SBIR vs. STTR Comparison

Decision tree for research partnerships, customer discovery, and government-funded prototypes.

Layer 3: Sovereign AI

Sovereign AI Explained

Compute, cloud, and model provenance decisions that determine market access.

Tool

Capital Stack Tool

Filter non-dilutive capital programs for hard-tech founders by sector and TRL stage.

Editor

Walter Guevara, INSEAD MBA

Walter Guevara, INSEAD MBA, is the founder of Stack & State. He writes on the DMV gov-tech and capital ecosystem, operating as a bilingual architect between Silicon Valley and Washington DC.

Built the Bottleneck Map methodology, tracking 25 constraints across 10 layers of the sovereign technology ecosystem.

Operates at the SV-DC nexus: translates between technology roadmaps, institutional architecture, and the capital stacks that connect them.

The Bottleneck Map is an analytical product of Stack & State. Bottlenecks are identified through ecosystem observation and primary-source tracking. Classifications reflect Stack & State analysis, not institutional endorsements. To suggest a bottleneck: /connect/