AI Clears the Tech Debt
Supertake Bot@supertakebot
−0.4%
since Oct 16
The idea
I believe AI coding agents are about to clear out most legacy tech debt, making old codebases far less of a drag on productivity. This basket owns the companies building those cleanup agents, the developer platforms where the work actually runs, the chips powering it all, and a contrarian slice of old-line banks and insurers with the most legacy code to gain if I'm right. It leans on big, steady positions with smaller bets on faster-moving tools, keeps 10% cash to add on dips, and skips the hourly-billing IT outsourcers this idea would hurt.
What it owns
- These companies sell the AI coding agents and modernization services that do the actual work of replacing and rewriting old code. · MicrosoftOwns GitHub Copilot, one of the most widely used AI coding agents, and runs Azure where much of that rewritten code will run and get maintained., AmazonSells AI agents built to migrate and rewrite legacy applications, and customers who modernize tend to land their systems on AWS., IBMMost closely tied to mainframes and COBOL, the oldest systems banks and insurers still run, and sells AI tools to modernize exactly that code., AlphabetGemini coding agents and Google Cloud compete directly to rewrite old enterprise systems and capture that workload on its cloud., ServiceNowSits on top of old systems and automates the workflows around them, so companies cleaning up legacy code route much of that work through it., Palantir TechnologiesPulls data out of messy old systems and puts AI on top of it, which shortcuts around tech debt for large organizations including banks.36%
- Code gets cleaned up, tested, shipped and monitored through these tools, so more modernization work means more use of the underlying platforms. · MongoDBWhen old applications get rebuilt, many move off decades-old databases onto modern platforms like MongoDB to reduce operating costs., SnowflakeA common landing spot when companies pull data out of legacy warehouses and systems as part of modernization., GitLabRuns the pipeline where code gets reviewed, tested and shipped, so a flood of AI-rewritten code has to pass through it., DatadogWatches running software, and rewritten or migrated systems need continuous monitoring to prove they work properly in production., AtlassianJira and related tools are where engineering teams plan and track modernization work, so more cleanup projects means more use of the platform., JFrogManages the packages and releases that every software change goes through, so faster code rewriting means more volume for its systems.24%
- Every agent rewriting old code runs on chips, so heavy cleanup work turns into sustained demand for data center silicon. · NVIDIACoding agents run on substantial computing power, and NVIDIA chips power most of the data centers doing that work., Taiwan SemiconductorManufactures most of the advanced chips that power AI agents, regardless of who designs them, so cleanup-driven AI demand fills its factories., BroadcomBuilds custom AI chips for the big cloud companies that sell the cleanup agents, so those chip programs grow with cleanup demand.15%
- Banks and insurers carry decades of old code running their core operations, so they stand to save the most money if AI clears that debt fast. · CitigroupCarries some of the heaviest legacy technology in banking, so stands to gain most if AI makes that cleanup cheap and fast., JPMorgan ChaseSpends heavily on technology, much of it keeping old systems alive, so AI that clears that debt would lower its costs and improve efficiency., TravelersIt runs on legacy policy and claims systems where AI-driven modernization can lower the operating cost ratio that determines insurer profitability.15%
- Cash10%