Six proprietary methodologies that structure how INA identifies, evaluates, finances and monitors digital infrastructure projects.
Each INA Framework™ works as a standalone methodology, but reaches its full value applied together, across a project's life — from first opportunity screen to post-implementation monitoring. Every framework is designed to evolve into a software module or AI-assisted capability as the INA platform matures.
INA's proprietary methodology for structuring, governing and implementing complex communications and data infrastructure projects — submarine cable systems, backbone and last-mile fiber, Fixed Wireless Access (FWA) deployments, and datacenters built for AI workloads. It is designed to apply in the same way to a public agency, a private company, an operator consortium, or a mixed-capital investment vehicle, and incorporates generative AI agents as a constitutive part of its architecture — one specialized agent per phase — always under a permanent human-in-the-loop principle.
“An infrastructure project doesn't fail at implementation: it fails at structuring. The INA Framework exists to fix that before it happens, regardless of the sponsor's nature.”INA Project Structuring Framework™ — Executive Summary
One specialized agent per phase, plus a cross-cutting traceability agent. None holds gate-approval authority — that RACI "Accountable" role always stays with a human.
| AI Agent | Phase | Primary Function | Human Checkpoint |
|---|---|---|---|
| Diagnostic Agent | Phase I | Analyzes the sponsor's existing documentation (contracts, network inventories, meeting records, prior surveys) and drafts the Strategic Diagnosis Document. | Executive sponsor validates before Gate 1 |
| Financial Structuring Agent | Phase II | Cross-references project parameters against financing lines (IDB, CAF, universal service funds, others) and generates a preliminary total-cost model. | Steering committee ratifies before Gate 2 |
| Contract Structuring Copilot | Phase III | Drafts technical specifications and tender documentation from INA's clause library, and flags atypical or high-risk clauses. | Legal counsel & technical committee validate before Gate 3 |
| Bid Evaluation Agent | Phase IV | Pre-processes and scores technical bids, and summarizes vendor progress reports in natural language. | Evaluation committee & PMO validate before award |
| Operations Intelligence Agent | Phase V | Continuously monitors KPIs and SLAs, detects anomalies, and drafts closure and lessons-learned reports. | PMO & executive sponsor validate before distribution |
| Traceability Agent | Cross-cutting | Maintains an automatic record of document versions, decisions and approvals tied to each gate. | None — audit record, not a decision function |
No AI agent in the Framework has authority to approve a gate, award a contract, commit funds, or accept a deliverable. Its role is always assistive: generating a first draft, flagging anomalies, or speeding up analysis that is then reviewed and approved by an accountable person.
Establish the problem to be solved, its alignment with the sponsor's strategic objectives, and the level of executive sponsorship available.
Determine technical, financial and regulatory viability, and build the business case that supports the investment decision.
Define how the project will be governed during execution and which contractual model transfers risk to vendors.
Execute the procurement process per applicable regulations and drive implementation with active control of scope, time, cost and quality.
Ensure operational adoption, sustain performance over time, and capture lessons learned for future projects.
The optimal combination depends on the sponsor's nature, the infrastructure typology, and the certainty of future revenues. A single project can combine more than one source.
| Financing Source | Instrument Type | Typical Use Cases | Key Considerations |
|---|---|---|---|
| Multilateral development banks (IDB, CAF, World Bank, FONPLATA) | Sovereign or sub-sovereign loans, concessional facilities | Submarine cable, national/regional backbone, large-scale rural connectivity | Requires sovereign backing/guarantee, safeguards, and the bank's procurement rules; 12–24 month origination |
| Universal service funds (e.g., Argentina's Fondo de Servicio Universal, administered by ENACOM) | Non-reimbursable grants, supply- or demand-side subsidies | Fiber in non-profitable localities, FWA in low-density areas, rural last mile | Subject to calls for proposals and fund regulations; requires demonstrating the connectivity gap and viability |
| National or sub-national public budget | Direct capital allocation | Modernization of existing public operator network, smaller-scale projects | Subject to the annual budget cycle and the public investment system |
| Private equity / infrastructure funds | Equity and quasi-equity | Datacenters (especially AI), FWA deployments by private operators | Requires a bankable business case, predictable cash flows and robust governance |
| Bank debt / project finance | Senior debt structured against project cash flows | Large-scale datacenters, submarine cable consortia | Requires solid revenue contracts (IRUs, colocation) and thorough technical due diligence |
| Vendor financing / export credit agencies (ECAs) | Supplier credit tied to specific technology acquisition | Cable-laying, network equipment, datacenter components | Requires vendor negotiation and compliance with the ECA's country-of-origin rules |
| Blended finance / Public-Private Partnership | Public risk mitigation (guarantees, viability-gap funding) combined with private capital | Rural fiber or FWA where standalone commercial returns are insufficient | Requires structuring the "viability gap" and a legal PPP framework in place |
| Bilateral development finance & trade agencies (USTDA, DFC) | Feasibility study grants (USTDA); direct loans, equity and political risk insurance (DFC) | Early-stage project preparation and long-term financing for U.S.-linked digital infrastructure | USTDA funds preparation only, not construction; DFC financing typically requires a U.S. nexus and a bankable structure |
The sponsor's highest authority, the executive sponsor, and user-area representatives. Approves gates and investment decisions.
Responsible for operational coordination and control of schedule, budget and risk.
Sponsor specialists and, where applicable, external INA advisors; validates architecture decisions and technical specifications.
R: Responsible · A: Accountable · C: Consulted · I: Informed. AI agents can generate drafts (R*) but never hold the Accountable role.
| Activity | Steering Committee | PMO | Technical Committee | Vendor | AI Agents |
|---|---|---|---|---|---|
| Gate approval | R/A | C | C | I | I |
| Schedule and budget management | I | R/A | C | C | C |
| Technical specifications | I | C | R/A | C | R* |
| Contract execution | I | A | C | R | C |
| Organizational change management | A | R | C | I | C |
| Risk | Probability | Impact | Primary Mitigation |
|---|---|---|---|
| Underestimation of migration costs | Medium | High | Total-cost model validated in Phase II with external benchmarking |
| Delays in the vendor selection process | High | Medium | Tender/RFP documentation legally pre-approved before publishing the call |
| Single-vendor dependency (vendor lock-in) | Medium | High | Contractual portability clauses and code/data ownership provisions |
| Low adoption by internal users | Medium | High | Change management plan starting in Phase III, not after delivery |
| Vendor SLA non-compliance | Medium | Medium | Escalating penalties and measurable indicators built into the tender |
| Regulatory changes during execution | Low | High | Review clause and technical committee with regulatory monitoring |
| Errors or inaccuracies in AI agent-generated content ("hallucinations") | Medium | High | No AI agent output is final without explicit human validation at the corresponding gate |
| Use of the sponsor's sensitive data in AI models without adequate controls | Medium | High | Data-handling policy defined in Phase III; controlled-retention environments and confidentiality agreements with AI providers |
Capital-intensive, 20–25 year lifecycle. Phase II is critical for structuring landing-party consortia, IRU capacity agreements and long-term financing. Phase III must cover repair clauses, landing permits per jurisdiction, and maritime/environmental coordination.
Often executed in geographic phases: each segment can run its own Phase II–IV cycle while Phase I and governance are defined once at the program level. Dominant risks: rights-of-way, municipal permits, utility coordination.
Structures faster than fiber or submarine cable. Phase II focuses on spectrum availability and cost; Phase III on CPE acquisition and deployment at scale. Lighter gates and short, iterative Phase IV cycles are recommended.
Particular power (high-capacity supply, N+1/2N redundancy), cooling and rack density requirements. Phase II must treat power availability and cost as a central business-case variable; Phase III must define GPU/accelerator supply agreements. Phase V must monitor energy efficiency (PUE) and availability.
Sourced from the INA Project Structuring Framework™ methodology document, v1.0 (July 2026).
A composite index that scores a project from 0–100 across eight weighted dimensions, translating readiness into a single, comparable number that investors and DFIs can act on quickly.
F2 is the natural first candidate for an interactive self-assessment tool: sponsors answer a structured questionnaire and receive an instant score with a prioritized improvement roadmap.
Evaluates a country, region or organization across ten domains of digital infrastructure maturity, from Initial to Leading, enabling governments and DFIs to benchmark progress and prioritize investment.
F3 is designed to become a live, continuously-updated country dashboard, combining public datasets with INA advisory inputs into an always-current maturity heat map.
Guides governments, DFIs and operators toward responsible, strategic AI adoption — across five pillars that move from ambition to governed, monitored deployment.
F4 is INA's own methodology for building AI capability — it will directly shape the INA platform's own AI advisory agents, dashboards and diagnostic tools as they launch.
Identifies, quantifies, prioritizes and monitors risk across ten categories throughout the project lifecycle — from strategic misalignment to climate exposure.
F5's risk register is a natural fit for automated monitoring: as project data updates, an AI layer can flag emerging risk before it becomes a stage-gate blocker.
Reads a project's country, sector, size, maturity and risk profile, then recommends which financing mechanisms — and which institutions — are the realistic fit.
F6 is the most natural candidate for a true decision-engine product: a structured intake that returns a ranked shortlist of financing mechanisms and matching institutions.