Answers for the technical evaluator.

Both — and more. The platform delivers a complete, production-ready system: application code, infrastructure-as-code (Terraform + Helm charts), a fully configured CI/CD pipeline, security hardening, API contracts, and documentation. The output isn't a starting point you hand off to engineers to finish — it's a deployable, operable system from day one.

Significantly faster than traditional delivery. The platform runs agent teams in parallel across architecture, code generation, security, and infrastructure — compressing what typically takes months into weeks. Exact timelines depend on scope and complexity, but the primary acceleration comes from eliminating sequential handoffs between design, development, QA, and DevSecOps.

The default stack is Java / Spring Boot, React, MySQL, Kubernetes, Helm, Terraform, and Jenkins. It's deliberately opinionated so that quality gates, security scanning, and operational tooling all work end-to-end without configuration overhead. The output is entirely standard, open-source tooling — no proprietary lock-in, no platform-specific runtime dependencies. Your team can take the output and run it independently on day one.

The platform is involved end-to-end — from requirements through to post-deployment operations. Code generation is one phase in a longer pipeline. After go-live, Sentinel AI continues operating the application: monitoring performance, investigating anomalies, automating resolution, and optimising over time. You're not handed a codebase and left to figure out operations independently.

The platform generates a fully configured Jenkins pipeline with all security and quality gates wired in. If your organisation uses a different CI/CD tool, the generated pipeline stages, scripts, and gate logic can be adapted — the underlying build artefacts and container images are standard OCI-compliant images that plug into most enterprise toolchains including GitHub Actions, GitLab CI, and Azure DevOps.

AWS, Azure, and GCP are all supported out of the box. The infrastructure layer is built on Terraform, so deployment targets are configurable rather than hardcoded. Private cloud and on-premise deployments are supported for organisations with data residency requirements, network isolation policies, or existing data centre investments you need to work within.

The Legacy Modernisation pipeline begins with an Understand phase where agents ingest your existing schema, data models, stored procedures, and documentation. This produces a structured map of your current data landscape before any new models are generated — avoiding manual re-mapping and significantly reducing the risk of data loss or logic gaps during migration.

Post-deployment operations run through Sentinel AI, the platform's AI-native ops layer. It handles monitoring, alert management, incident investigation, and resolution through natural language — no separate ops tooling or war-room process required. Your team retains full ownership and visibility; Sentinel augments rather than replaces your engineers, escalating to them with root cause context when autonomous resolution isn't possible.

Sentinel runs a continuous Observe → Investigate → Act → Optimize loop. It detects anomalies, correlates signals across services, identifies root cause, and attempts automated resolution. If it can't resolve autonomously, it escalates to your team — not with a raw alert, but with a full root cause analysis and a suggested fix already in hand.

The generated codebase is fully yours — standard, documented, and readable by any competent engineering team. There are no proprietary hooks, platform-specific runtime dependencies, or lock-in patterns. The walk-away principle means your team can take the output and operate it entirely independently if they choose. Post-modification, Sentinel AI continues monitoring the deployed application at the infrastructure and service level, regardless of what changed in source.