AIM

1. Start with a Guided Assessment

AIM begins with a structured intake that captures your environment in plain language. You answer questions — AIM normalizes and structures the rest.

Scope-conditional sub-questions automatically appear based on what you select — so you're only ever asked what's relevant to your project.

2. Constraint Normalizer (C-1, C-2, C-3…)

Free-text constraints are converted into normalized constraint records with stable IDs:

Each constraint is split into atomic statements, classified by type, and given a stable code. These codes appear in every recommendation and report so decision-makers can see exactly which constraints each recommendation honors.

3. Free-Text Signal Extraction

When you describe your constraints in free text, AIM automatically scans the language for structured signals — deterministically, at zero AI cost — and pre-fills relevant assessment fields without overwriting anything you've already provided.

Extracted signals are stored on the assessment and used to skip redundant follow-up questions. You can trigger re-extraction any time from the Smart Follow-Up tab.

4. Structured Context Builder

Behind the scenes, AIM assembles a structured context package — a curated, normalized representation of your situation that AI models are allowed to use. Raw survey answers are never passed directly to AI.

This structured context — not your raw inputs — is what AIM's AI models work within. The package is built deterministically before any AI is invoked.

5. Assessment Confidence & Smart Follow-Up

AIM measures how complete your assessment data is and assigns a confidence tier that appears on every generated report. The more signals you provide, the more precise the recommendations and the stronger the report language.

6. RAO Engine (Ranking and Optimization)

AIM doesn't just “ask a model for ideas.” RAO is a scoring system that ranks modernization paths using your constraints as hard boundaries — not suggestions. It evaluates options across multiple dimensions:

• Cost – upfront + operating
• Maturity – stability and real-world adoption
• Security & compliance – fit against your regulatory profile
• Vendor lock-in risk – difficulty of future exit
• Operational complexity – skills, runbooks, and support needs
• Strategic fit – alignment with constraints C-1, C-2, C-3, etc.

Each recommendation carries a RAO score and a dimension breakdown — so you can explain why a path is preferred, not just what the AI suggested.

7. Risk & Integration Insight Engines

AIM analyzes your inputs to surface:

8. Anti-Hallucination Guardrails

AIM is deliberately strict about what it's allowed to say. The report and recommendation engines are explicitly instructed to:

• Use only facts present in the structured context
• Never invent vendors, technologies, integrations, or regulations
• Recommend named technologies only when supported by your inputs — and flag anything net-new
• Only mention HIPAA, CJIS, PCI, etc. if detected in your constraints or requirements
• Tie recommendations back to RAO scores and constraint codes

That's by design. Your assessment confidence tier documents exactly how much data was provided and what was assumed.

9. Intelligence Layers

Beyond the core RAO engine, every assessment passes through a set of proprietary intelligence layers that enrich outputs deterministically — no AI guesswork:

• ›AI Readiness & Archetype Classification — When AI is in scope, AIM scores readiness and classifies the appropriate AI implementation approach.
• ›Modernization Focus Classification — Automatically infers the primary modernization archetype from assessment signals, adjusting RAO scoring weights downstream.
• ›FTE Estimation Engine — Produces a staffing range and duration estimate per archetype, broken out by delivery role.
• ›Labor Cost Intelligence — Computes delivery cost ranges across three models (market hiring, federal staffing, contractor) using authoritative federal rate data.
• ›Historical Procurement Benchmarking — Compares your project against real public-sector contract award data to surface comparable cost and duration ranges.
• ›Sustainment Cost Analysis — When enabled, extends the cost picture through long-term operational budgeting, staffing gap analysis, and 3-year TCO projections.

10. Outputs: What AIM Actually Delivers

From a single assessment, AIM can generate:

11. Why AIM Is Different

• Objective by design – AIM is not paid by vendors, marketplaces, or cloud providers.
• Constraint-driven – C-codes make your priorities explicit and traceable.
• Transparent – Recommendations and reports show why, not just what.
• Confidence-aware – Every report documents exactly what was provided and what was estimated.
• Built for government & regulated environments – Healthcare, public safety, and financial contexts are first-class citizens.

12. Teams & Collaboration

AIM supports a multi-stakeholder model with six organization roles (Owner, Admin, Engineer, Program Analyst, Reviewer, Viewer) and a separate guest access system for external participants.

See the full Teams & Access guide →

13. Product Catalog Query

Every assessment workspace includes a Product Catalog Query Widget that lets users investigate the technology catalog directly — without leaving the assessment.

14. Technology Catalog

AIM recommendations draw from a curated catalog of 250+ enterprise-grade technology products across 26 categories — including ERP, Identity & Access, Observability, Data Warehouse, ETL/Integration, and a full AI/ML suite covering Platforms, Generative AI, Data Science, MLOps, NLP, Computer Vision, AI Ethics, and AI Security. AI tools are compliance-filtered against CJIS, FedRAMP, HIPAA, and PCI based on your declared regulatory scope.