# MODIFIABLE REQUEST FOR PROPOSALS (RFP) TEMPLATE
## Project Title: PowerBuilder to Google Cloud Migration (Cloud Run, React & AlloyDB)
- **RFP Document Reference**: `VNA-RFP-002` (Companion to Solution Brief `VNA-SB-002`)
- **Issuing Organization**: `[INSERT ORGANIZATION / COMPANY NAME]`
- **Issue Date**: `[2026-10-27]`
- **Proposal Due Date**: `[INSERT DATE, e.g., +30 Calendar Days]`
- **Reference Architecture Portal**: https://powerbuilder-to-gcp.web.app/
- **Template Attribution**: This site is designed by Viscosity.AI

---

## SECTION 1: ISSUING ORGANIZATION PROFILE & PROCUREMENT SCHEDULE (MODIFIABLE)

> **Instructions for RFP Author**: Replace all bracketed placeholders (`[...]`) below with your organization's specific dates, contacts, and environment metrics before issuing to prospective bidders.

| Procurement Milestone | Target Date / Entry (Modifiable) |
|---|---|
| Issuing Organization Name | `[Insert Legal Entity Name]` |
| Primary Procurement / Technical Contact | `[Insert Name, Title, Email]` |
| RFP Issue Date | `[2026-10-27]` |
| Vendor Written Questions Due | `[Insert Date - Week 2]` |
| Organization Responses to Vendor Q&A | `[Insert Date - Week 3]` |
| Final Vendor Proposals & Pricing Due | `[Insert Date - Week 4]` |
| Technical Architecture Demos / Oral Presentations | `[Insert Date - Week 5]` |
| Target Contract Award & Phase 0 Kickoff | `[Insert Date - Week 6]` |

### 1.1 Executive Summary & Business Objectives
`[INSERT COMPANY NAME]` ("Organization") is soliciting competitive proposals from qualified enterprise architecture and engineering firms to modernize and transition our current **PowerBuilder (.pbl, DataWindows, PowerScript) + Sybase / SQL Server / Oracle** environment to **Google Cloud Run, React/TypeScript, AlloyDB / Cloud SQL & BigQuery**.

Our primary business and technical objectives are:
- Organizations standardizing on Google Cloud Platform often discover dozens of departmental or core operational PowerBuilder applications tied to on-premises Sybase ASE, SQL Server, or legacy Oracle instances.
- Lifting and shifting Windows fat-client binaries into Google Compute Engine VMs with remote desktop streaming preserves every maintenance bottleneck while inflating cloud compute bills.
- Engineering teams proficient in TypeScript, Python, and Go lack the domain tooling to decipher binary PowerBuilder .pbl libraries, inherited DataWindow buffers (Primary!, Delete!, Filter!), and implicit transaction commits.

---

## SECTION 2: CURRENT ENVIRONMENT BASELINE INVENTORY (MODIFIABLE)

Bidders must size their technical delivery plan and fixed-fee milestone pricing against the baseline inventory table below:

| Inventory Dimension | Current Baseline Count / Footprint (Modifiable) | Target Modernization SLA / Objective |
|---|---|---|
| Legacy Source Stack (`Stack X`) | PowerBuilder (.pbl, DataWindows, PowerScript) + Sybase / SQL Server / Oracle | Full retirement / modernization |
| Target Architecture (`Stack Y`) | Google Cloud Run, React/TypeScript, AlloyDB / Cloud SQL & BigQuery | Production-grade HA/DR & security |
| PowerBuilder Windows & Services | `[Insert Count, e.g., 180 windows]` | 100% functional parity & optimization |
| DataWindows & Stored Procedures | `[Insert Count, e.g., 400 objects]` | Automated AST & schema uplift |
| Query Fan-Out Rationalization (Cloned DataWindows -> Core SQL + Runtime Filters) | `[Insert Count, e.g., 75 % fan-out reduction]` | `[Target Rationalization / SLA]` |
| Target Cutover Downtime Window | `[e.g., < 15 Minutes / Near-Zero Downtime]` | Rehearsed parallel run + rollback plan |
| Security, Compliance & Identity | `[e.g., SSO/MFA via Entra ID / Okta, Row-Level VPD, Audit Logs]` | Zero critical/high CVEs at Go-Live |

---

## SECTION 3: MANDATORY TECHNICAL SCOPE OF WORK & DELIVERABLES (MODIFIABLE)

The selected partner shall execute the project across four gated phases (**Phase 0: AST Discovery & Architecture Blueprint**, **Phase 1: Core Foundation & Data/Code Conversion**, **Phase 2: Integration, Runtime Filtering & UAT**, and **Phase 3: Parallel Run, Cutover & Hypercare**).

### 3.1 Core Architectural Requirements
1. **[MODIFIABLE REQUIREMENT 1]**: Automated PBL Decompilation & Specification Blueprinting: Our reverse-engineering engine parses every .pbl library, producing OpenAPI 3.1 specifications, Entity-Relationship models, and human-readable business-rule specifications from legacy PowerScript.
2. **[MODIFIABLE REQUIREMENT 2]**: DataWindow Buffer-to-REST/GraphQL Translation: PowerBuilder's stateful Primary!, Filter!, and Delete! DataWindow buffers are translated into stateless, optimistic-concurrency REST endpoints running on Google Cloud Run (Python FastAPI / Node / Go) backed by Cloud Tasks and Pub/Sub.
3. **[MODIFIABLE REQUIREMENT 3]**: Database Modernization to AlloyDB or Oracle@Google Cloud: Underlying Sybase, SQL Server, or legacy Oracle schemas are migrated via Google Database Migration Service (DMS) or GoldenGate into AlloyDB for PostgreSQL or Oracle Database@Google Cloud.
4. **[MODIFIABLE REQUIREMENT 4]**: Reporting DataWindows to BigQuery & Looker: Heavy reporting and composite DataWindows are offloaded from transactional tables into BigQuery medallion tables with scheduled queries and Looker Studio dashboards.
5. **[MODIFIABLE REQUIREMENT 5]**: Query Fan-Out Rationalization & Runtime/Inference Filtering: Instead of migrating cloned reports and DataWindows 1-for-1, our AST SQL parser clusters queries by their canonical core JOIN graph, builds governed Oracle 23ai views only for those core queries, and generates dynamic WHERE clauses, faceted filters, and ad-hoc report variations at runtime and via AI inference (APEX 24.2 Smart Filters & Select AI).

### 3.2 Component-Level Conversion & Deliverable Matrix
Bidders must confirm compliance with each component conversion requirement below:

| Req ID | Legacy Artifact / Baseline Workload | Required Target Architecture | Engineering Method & Acceptance Criteria | Priority (Modifiable) |
|---|---|---|---|---|
| REQ-01 | PowerBuilder DataWindow (.srd) CRUD & Buffers | React TanStack Table / AG Grid + FastAPI on Cloud Run | OpenAPI 3.1 Schema & JSON DTO Generation | `[Mandatory / Optional]` |
| REQ-02 | PowerScript Non-Visual Objects (NVO) | Containerized Cloud Run Microservices (Python / TypeScript) | AST Business Rule Extraction + Pytest Suite | `[Mandatory / Optional]` |
| REQ-03 | Sybase ASE / SQL Server Stored Procedures | AlloyDB for PostgreSQL PL/pgSQL or Oracle@GCP | Automated Dialect Transpilation & DMS Cutover | `[Mandatory / Optional]` |
| REQ-04 | Composite Reporting DataWindows | BigQuery Partitioned Views + Looker / PDF Cloud Functions | SQL Pushdown & Serverless PDF Rendering | `[Mandatory / Optional]` |
| REQ-05 | INI Files & Windows Registry Config | GCP Secret Manager + Cloud Identity / IAP SSO | Zero-Trust IAM & OAuth2 Token Wiring | `[Mandatory / Optional]` |
| REQ-06 | Cloned Reports / Queries Differing Only by WHERE Clause or Group-By (Query Fan-Out) | Canonical Core SQL View + APEX 24.2 Runtime / AI Inference Filters | AST Join-Graph Clustering (75-90% Report Count Reduction) | `[Mandatory / Optional]` |

### 3.3 Mandatory Platform & Tooling Standards
- Viscosity PBL/SRD AST Reverse-Engineering Parser & Blueprint Generator
- Google Cloud Run, Cloud Build, Artifact Registry, and Identity-Aware Proxy (IAP)
- AlloyDB for PostgreSQL, Oracle Database@Google Cloud, and BigQuery
- Oracle GoldenGate / Google Database Migration Service (DMS) for zero-downtime cutover
- Vertex AI & Gemini Code Assist for automated unit-test synthesis

---

## SECTION 4: VENDOR TECHNICAL Q&A QUESTIONNAIRE & SCORING RUBRIC

Bidders must provide direct, concise written responses to each question below. The Issuing Organization's evaluation committee will grade responses using the embedded benchmark criteria.

### Vendor Question 1: Can we run Oracle Database natively inside Google Cloud if we want to keep our PL/SQL packages?
- **Bidder Response Requirement**: Provide a concrete architectural explanation, reference toolset, and verifiable production proof. Generic marketing statements will be scored zero.
- **Evaluation Benchmark (What High-Scoring Responses Must Demonstrate)**: Yes. With Oracle Database@Google Cloud (Exadata Database Service and Autonomous Database running physically inside GCP data centers), you can pair Google Cloud Run / Vertex AI directly with low-latency Oracle databases or Oracle APEX.
- **Evaluator Score (0 to 10 Points)**: `[   / 10 ]`

### Vendor Question 2: How do you validate that the new Google Cloud application matches legacy PowerBuilder calculations?
- **Bidder Response Requirement**: Provide a concrete architectural explanation, reference toolset, and verifiable production proof. Generic marketing statements will be scored zero.
- **Evaluation Benchmark (What High-Scoring Responses Must Demonstrate)**: Our pipeline captures real transaction payloads and DataWindow result sets from the legacy app and replays them against the new Cloud Run API endpoints to prove 100% numerical parity before cutover.
- **Evaluator Score (0 to 10 Points)**: `[   / 10 ]`

### Vendor Question 3: What is included in the downloadable PDF Solution Brief?
- **Bidder Response Requirement**: Provide a concrete architectural explanation, reference toolset, and verifiable production proof. Generic marketing statements will be scored zero.
- **Evaluation Benchmark (What High-Scoring Responses Must Demonstrate)**: The PDF Solution Brief details the 5-stage PowerBuilder-to-GCP AST extraction pipeline, the DataWindow-to-React/Cloud-Run reference architecture, and a decision matrix comparing AlloyDB vs. Oracle Database@Google Cloud.
- **Evaluator Score (0 to 10 Points)**: `[   / 10 ]`

### Vendor Question 4: How do you prevent report and query fan-out bloat, and how do you guarantee zero vendor lock-in?
- **Bidder Response Requirement**: Detail your methodology for clustering redundant legacy queries/reports into canonical core SQL views with runtime or AI-inference filtering, and specify customer code/data ownership upon project completion.
- **Evaluation Benchmark (What High-Scoring Responses Must Demonstrate)**: Bidder must demonstrate deterministic AST SQL clustering (collapsing 75%-90% of cloned reports into core database views with runtime predicates), 100% customer ownership of all DDL, PL/SQL, application source code, and Terraform scripts, and zero proprietary runtime lock-in.
- **Evaluator Score (0 to 10 Points)**: `[   / 10 ]`

### Vendor Question 5: Describe your automated regression testing, parallel run, and zero-downtime cutover methodology.
- **Bidder Response Requirement**: Specify how data parity, business-rule equivalence, and rollback safety are verified prior to production cutover.
- **Evaluation Benchmark (What High-Scoring Responses Must Demonstrate)**: Bidder must provide automated row-count and cryptographic hash reconciliation, automated unit/UI test suites (e.g., utPLSQL, Playwright), and a rehearsed cutover runbook with sub-minute rollback capability.
- **Evaluator Score (0 to 10 Points)**: `[   / 10 ]`

---

## SECTION 5: COMMERCIAL PRICING & LICENSING MATRIX (MODIFIABLE)

Bidders must complete the fixed-price milestone table below and disclose all recurring software or cloud consumption assumptions.

| Milestone / Cost Category | Deliverable & Acceptance Gate | Fixed Fee / Annual Cost (USD) | Estimated Duration (Weeks) |
|---|---|---|---|
| **Phase 0: Discovery & AST Code/Schema X-Ray** | Verified inventory, dead-code manifest, core SQL fan-out map, and final blueprint | `$[Insert]` | `[2 - 3 Weeks]` |
| **Phase 1: Target Environment & Automated Uplift** | Provisioned target stack, schema migration, and core business logic conversion | `$[Insert]` | `[4 - 8 Weeks]` |
| **Phase 2: UI, Integrations, Reporting & UAT** | Completed application/portal screens, runtime reports, AOP templates, and UAT sign-off | `$[Insert]` | `[3 - 6 Weeks]` |
| **Phase 3: Production Cutover & Hypercare** | Rehearsed cutover, reconciliation sign-off, documentation, and 30-day hypercare | `$[Insert]` | `[2 - 3 Weeks]` |
| **Optional: Annual Managed Services / Support** | 24x7 monitoring, patching, DBA performance tuning, and quarterly enhancements | `$[Insert / yr]` | `[12 Months]` |

---

## SECTION 6: PROPOSAL EVALUATION SCORECARD (MODIFIABLE)

| Evaluation Criterion | Weight (%) | Description |
|---|---|---|
| **1. Technical Architecture & Automation Tooling** | `35%` | Depth of AST automation, query fan-out rationalization, database/cloud architecture, and zero lock-in. |
| **2. Verifiable Production Experience & References** | `25%` | Proven track record delivering identical `PowerBuilder to Google Cloud Migration` engagements with measurable outcomes. |
| **3. Cutover Safety, Testing & Compliance** | `20%` | Rigor of data reconciliation, automated regression testing, security/compliance controls, and rollback safety. |
| **4. Commercial Value & Total Cost of Ownership (TCO)** | `20%` | Transparent fixed-fee milestone pricing and 3-to-5 year software/cloud licensing savings. |
| **TOTAL** | **`100%`** | |

---
*Template Reference: `VNA-RFP-002` | Companion Portal: https://powerbuilder-to-gcp.web.app/ | This site is designed by Viscosity.AI*