6-Week Windows-First POC Plan

POC objective

Give executive leadership a fast, hands-on view in a Windows-only POC environment: run one representative COBOL change and show clear before/after evidence for cycle time, delivery control, and AI-assisted workflow quality.

This follows the same approach used in the reference ~9-month Ingenium migration.

📧 Discuss your POC scope

What we set up in the POC

• Windows-only pilot setup for executive visibility and safe evaluation
• Nexus toolchain in VS Code for one Ingenium application area
• Automated COBOL build & deploy for a representative change
• One consistent, repeatable environment (removing manual setup drift)
• Optional: one REST API endpoint exposing an existing Ingenium function

Trial license packaged with the POC

• A 6-month trial license is included after the 6-week POC to extend hands-on usage beyond the pilot.
• Recommended governance model: checkpoint reviews at month 2 and month 4, then convert to production commercial terms only if KPI and governance gates are met.

AI enablement included in the POC package

This POC includes a practical AI working model based on the real Ingenium project setup in .github/agents and .github/skills.

Skill and agent inventory is managed as a living operating model. The list below is the current baseline at POC kickoff and can change over time based on tool maturity, governance requirements, and delivery scope.

1) Nexus skills baseline at POC kickoff

Skill	What it does in POC
icomp	COBOL compile workflows (quick, mass, golden, bind) for faster change cycles
igit	Work-item Git flow and source-point delta (info, diff, list, golden)
ing	Worker lifecycle control by region (start, stop, status)
iadm	DB2 administration operations for controlled environment setup
idb	Backup/restore runs to protect rollback and recovery confidence
nexus	Security utilities (encrypt, decrypt, hash, lic) for safe operational handling

2) Specialized agents baseline at POC kickoff

Agent	POC role	Persistent output
cobdoc	Technical analysis of work-item change	<WORK-ITEM-ID>_analysis.md
cobimpact	Impact matrix and regression scope tiers	<WORK-ITEM-ID>_impact.md
cobreview	Risk-first review with prioritized findings	<WORK-ITEM-ID>_review.md
cobtest	Requirement-to-test traceability and blockers	<WORK-ITEM-ID>_testplan.md
cobsuite	Orchestrates full pipeline across agents	Consolidated artifact set in doc/items/<branch>/

3) Why this matters for executive decisions

• AI support is tied to real delivery operations, not chat-only assistance.
• Each agent creates persistent, auditable evidence files for governance reviews.
• Go/no-go decisions at the end of POC are based on measurable output quality, cycle time, and risk visibility.

4) AI workflow used during POC

Execution blueprint

Success metrics

KPI	What we measure
Build & deploy time	Time from an approved COBOL change to a deployed, running build (before vs after)
Reliability	Failed/rolled-back deployments and environment inconsistencies
Manual effort	Engineer hours spent on build/deploy/environment setup
Auditability	Whether every release and environment change is traceable
AI evidence quality	Completeness and usability of generated _analysis, _impact, _review, _testplan artifacts for one real work item
Trial adoption readiness	Whether teams can continue with the included 6-month trial under agreed governance checkpoints

Governance cadence

• Weekly steering checkpoint with executive sponsor delegate and business owner
• Bi-weekly risk and dependency review
• End-of-POC evidence review with the transformation office / CFO delegate

Exit criteria

1. The build/deploy time improvement is clear and attributable
2. Risks are documented with mitigation path and residual impact
3. Leadership receives a prioritized, costed scale-out roadmap
4. Windows-first pilot is approved to continue under the 6-month trial license model

---

Ready to define your POC?

📧 Schedule discovery call
⏱️ Response within 2 business days
🎯 No-obligation consultation