--- title: "AI-Assisted Application Modernization | Smart Data" url: https://www.smartdata.net/services/ai-assisted-modernization description: "Engineers using Claude and current models to close vulnerabilities, upgrade Angular and .NET, consolidate codebases and convert legacy ETL to Databricks." source: "Smart Data (smartdata.net). Markdown copy of the page; the HTML page is canonical." --- AI # AI-assisted application modernization Our engineers use Claude and other current models on the slow, high-volume work in older estates: security backlogs, framework upgrades, codebase consolidation and legacy ETL. Here is what that looked like on two client engagements. “I attribute much of the quality of the work and especially the compressed timeline to Smart Data’s mastery of AI,” the client’s IT leader wrote. ## Customer portals secured and upgraded 620+ known vulnerabilities eliminated Two engagements, in detail ## Customer portals for a national distributor, legacy ETL for a national retailer Upgrades, remediation and conversions are where AI changes the work most: there is a lot of it, the details matter, and tests and scanners can confirm each change. Security remediation, framework upgrades and codebase consolidation come from the distributor engagement, and legacy ETL conversion comes from the retailer. Security remediation ### Vulnerability and security remediation Scanner findings, open penetration test items and the dependency debt behind them, closed across whole application portfolios. On one engagement for a national distributor, that meant more than 620 known vulnerabilities eliminated and every penetration test finding remediated, with results cross-checked against a second vulnerability database that surfaced issues the first scanner missed. - Commercial scanner findings remediated and cross-checked against the open-source OSV database - Penetration test findings closed, with the evidence your security team needs - Real vulnerabilities found beyond what the scanner reported, and fixed Tools we work in - OSV - OWASP Framework upgrades ### Framework and runtime upgrades Applications several major versions behind are a security problem before they are a feature problem. We bring frameworks and runtimes current: on the distributor’s portals, Angular 16 to 21 and Azure Functions from the .NET 6 in-process model to the .NET 10 isolated worker, with Angular bundles about 50 percent smaller. - Several major versions closed in one engagement - Packages updated across every project - Dead infrastructure found and removed along the way: more than $25K a year realized on that engagement - Angular - .NET - Azure Functions - TypeScript Codebase consolidation ### Codebase consolidation and multi-tenancy Many estates carry the same application deployed once per customer or business unit, each copy drifting and each one billed. We consolidate them into one multi-tenant set of services, so a new customer means a new database rather than a new suite of cloud resources. For the distributor, five to six copies of one codebase are moving onto one multi-tenant platform, with about $78K a year in projected savings. - Per-customer copies mapped, compared and folded into shared services - A new customer becomes a new database, not a new suite of Azure resources - Duplicated cloud resources retired as each tenant moves across - Azure - Multi-tenant architecture Legacy ETL conversion ### Legacy ETL conversion to Databricks For a major national retailer we built an AI-assisted migration tool, using Claude and structured prompts, that converts legacy ETL logic, orchestration scripts and deployment configurations into PySpark pipelines on Databricks. Unit tests and infrastructure as code are generated with every conversion, and in the first phase the tool handled 78% of the legacy transformation types. - Legacy transformations converted into PySpark pipelines on Databricks - Unit tests and Terraform generated alongside every pipeline - Senior engineers validating output against known behavior instead of translating by hand - Databricks - PySpark - Claude - Terraform - GitHub Actions The retailer migration tool ## How the tool turns legacy ETL into Databricks pipelines The retailer had hundreds of legacy workflows and orchestration jobs. We built a tool on Claude and structured prompts instead, and tested it across a real cross-section of the estate in a three-month first phase. ### 01Scope from real production 31 workflows, 36 mappings, 60 orchestration jobs and 50 shell scripts, drawn from live domains rather than a curated sample. ### 02Convert with Claude Legacy ETL logic, orchestration scripts and deployment configuration converted into PySpark pipelines on Databricks. ### 03Generate the proof Unit tests, plus Terraform and GitHub Actions deployment code, generated alongside every converted pipeline. ### 04Validate, then run Senior engineers validate output against known behavior instead of translating by hand, and the platform team can run what they review. Industries ## Industries we work in ### Healthcare Patient, pharmacy and clinical data systems held to HIPAA, including where AI is allowed to touch them. - Medicaid programs - Pharmacy services - Clinical documentation [Healthcare](https://www.smartdata.net/industries/healthcare) ### Manufacturing and industrial ERP, shop-floor and supplier data on one platform, so plant reporting and finance reporting reconcile. - SAP and ERP reporting - Shop-floor integration - Supplier portals [Manufacturing and industrial](https://www.smartdata.net/industries/manufacturing) ### Retail and distribution Order, inventory, credit and customer data connected across portals, branches and the platforms behind them. - Vendor and customer portals - Order and credit systems - Data platforms [Retail and distribution](https://www.smartdata.net/industries/retail) ### Financial services Bank data platforms, reporting that holds up to an audit and managed support for the financial systems the business runs on. - Bank data platforms - Managed ERP support - Legacy migration [Financial services](https://www.smartdata.net/industries/banking-financial-services) ### Insurance Claims, member and policy systems for health plans and insurers. - Health plans - Claims data - Member apps [Insurance](https://www.smartdata.net/industries/insurance) Case studies ## Modernization on client work ### known vulnerabilities eliminated across the customer portals One AI-assisted senior engineer closed every known vulnerability ahead of plan, then started consolidating five to six copies of the codebase into one set of services. About $78K a year projected, over $25K a year already realized. - AI-Assisted Development - .NET Core [AI-assisted delivery cleared 620+ vulnerabilities, then rebuilt the platform: known vulnerabilities eliminated across th](https://www.smartdata.net/case-studies/application-security-remediation-multi-tenant-consolidation) 78% ### of legacy transformation types handled by the AI-assisted migration tool An AI-assisted tool that converts legacy ETL into tested Databricks pipelines. In a three-month first phase across 177 legacy jobs and scripts, it handled 78% of the legacy transformation types, with tests generated alongside every conversion. - ETL & Data Pipelines [Legacy ETL to Databricks for a major national retailer: of legacy transformation types handled by the AI-assisted migrat](https://www.smartdata.net/case-studies/databricks-migration-national-retailer) ### From green-screen to modern web for one of America's largest industrial distributors Winsupply’s IBM RPG order management, moved to an Angular and Java web platform where customers manage their own orders and see their supply chain. - Java - Spring Boot [From green-screen to modern web for one of America's largest industrial distributors](https://www.smartdata.net/case-studies/winsupply-order-management-modernization) A few practical questions ## Questions about AI-assisted modernization What engineering leaders usually ask before the first codebase is opened. Mostly Claude, with other current models where they suit the task better. The retailer migration tool runs on Claude with structured prompts. We only use AI tools with training on your data turned off, and if your contract restricts AI or requires stricter handling, your contract wins. On the distributor’s portals, every commercial scanner finding was closed, the results were cross-checked against the open-source OSV database, which surfaced real issues the scanner missed, and every penetration test finding was remediated. Yes. Both engagements on this page were on software someone else wrote, years earlier. Getting the current behavior pinned down, and writing tests where they are missing, comes first. Usually small and specific: one codebase and its backlog, or a first phase of a migration. The distributor engagement started with the security backlog and grew into the consolidation. No. Those are the stacks on these two engagements. The same approach works on other stacks, and our modernization work from before today’s AI tools reaches IBM i and VB6 desktop software. Related services ## Where modernization connects ### AI on your data Governed, private AI that answers questions from your own data, with the query behind every answer. [AI on your data](https://www.smartdata.net/services/ai-on-your-data) ### AI development Language model features built into the applications your people already use. [AI development](https://www.smartdata.net/services/ai-development) ### Application development and modernization Custom applications, portals and cloud-native rebuilds of the software that runs the business. [Application development and modernization](https://www.smartdata.net/services/web-solutions) ### Databricks consulting Lakehouse architecture and legacy ETL migration with AI-assisted conversion and generated tests. [Databricks consulting](https://www.smartdata.net/services/databricks-consulting) Next step ## Which codebase would you modernize first? Tell us what it runs on and what is holding it back. Here is what happens after you reach out. ### A short call about the codebase What it runs on, what the scanners say and what is driving the timing. ### An engineer looks at what’s running Scope comes from your systems as they are, before anything is priced. ### You get a written proposal Covering scope, team, approach and price. [info@smartdata.net](mailto:info@smartdata.net)[(937) 886-9166](tel:+19378869166)