Infinite Innovation Newsletter
Two AI / Data Giants We Need To Own For The Long Term: Palantir & Databricks
Happy Tuesday night and welcome back to the Infinite Innovation Newsletter from Coach Kyle and G.L.A INVE$T!!
I get asked a lot, who’s the next PALANTIR? Well, that’s easy! PALANTIR is the next PALANTIR!! and DATABRICKS is the Robin to $PLTR’s Batman. They are not “apps on top” of AI; they are becoming the operating fabric for how data is stored, governed, transformed, and turned into real‑world action at scale.
At G.L.A INVE$T RESEARCH, Palantir remains the second‑favorite long‑term public company, a name that belongs in the same sentence as past generational winners like Microsoft, while Databricks is firmly in the top three favorite private companies, with an IPO on the horizon that could reshape the public AI landscape. This issue breaks down why both are core long‑term “own, not trade” positions for serious innovation investors and why their partnership is one of the most important (and misunderstood) AI stories in the market right now.
Palantir: This Generation’s Microsoft Of Operational AI
Business and mission
Palantir builds full‑stack software platforms that let governments and enterprises integrate their messy real‑world data, reason over it, and act on it in live operations. The core stack is:
Foundry: data integration, modeling, and operational applications for commercial customers.
AIP (Artificial Intelligence Platform): the AI control plane that connects LLMs to an organization’s governed data and workflows, turning AI from a chatbot into an engine of decisions and actions.
Gotham & Apollo: platforms for defense/intel and software deployment at scale.
This is not a simple analytics or dashboard company. Foundry and AIP are designed so that the same platform that cleans and models your data also powers the applications your operators, analysts, and managers use to make decisions and write back to operational systems in real time.
Growth, founders, and achievements
Palantir was founded in 2003 by Alex Karp, Peter Thiel, and a small group of engineers focused on building software for counter‑terrorism and intelligence, with a design philosophy of “human‑driven, machine‑accelerated” decision‑making. Over the last few years, commercial has gone from “side business” to major growth engine, helped by AIP, which drastically shortens sales cycles by letting customers spin up high‑value AI workflows on their own data fast.
Among achievements that matter to long‑term investors:
Deeply embedded in critical Western government and defense infrastructure, often chosen where failure is not an option.
AIP adoption driving rapid expansion of use cases per customer, as enterprises realize they can move from POCs to production AI faster than with DIY stacks.
TAM, moat, and problems they solve
The addressable market is not just “analytics”; it’s the full stack of data integration, governance, AI model orchestration, and mission‑critical operational software across defense, manufacturing, healthcare, finance, and more ( Basically every company and every military ally will eventually use $PLTR and their platform)That spans hundreds of billions in enterprise IT and defense digital modernization spend over the next decade.
Palantir’s moat is structural, not cosmetic:
Ontology: a living map of an enterprise, modeling real‑world objects (assets, shipments, patients, missions) and their relationships so humans and AI can reason over them.
Multimodal security: fine‑grained access control designed for classified and highly regulated environments, now extended into the AI era via AIP’s “AI inside your perimeter” design.
End‑to‑end integration: data pipelines, business logic, UI, workflow engines, and AI all in one environment, which makes it very hard to rip out once embedded.
The problems Palantir solves: organizations are drowning in data, regulation, and complexity, with AI projects stuck as demos. Palantir turns that tangle into governed, operational AI systems that actually change how factories run, how logistics are optimized, and how defense missions are executed.
Why PLTR is a core long‑term position
PLTR remains one of our favorite long‑term public name on the G.L.A. list, I have owned the company since before they came public as a pre ipo. I view $PLTR as this generation’s MSFT: software operating layer for data and AI‑driven decision‑making across industries. Its AI platform is one of the best in the world for connecting LLMs to real, governed operational data, and the company is strategically critical for the United States and its allies. With Alex Karp and Peter Thiel as star founders who think in multi‑decade horizons, Palantir is disrupting legacy consultants, “dashboard‑only” BI tools, and fragmented data stacks rather than being disrupted by them. Bottom line, if a company wants to dominate, they hire Palantir.
Databricks: The Data & AI Engine Of The Enterprise
Business and evolution
Databricks started as the commercial home of Apache Spark, solving the pain of managing complex big‑data compute clusters and turning them into a managed platform. It has since evolved into a full Data Intelligence Platform that unifies:
Lakehouse storage (Delta Lake / Lakebase)
Data engineering and ETL (Lakeflow)
Warehousing and BI
Governance (Unity Catalog)
AI and agents (Mosaic AI, Agent Bricks, Genie)
The goal: be the place where all enterprise data lives, is governed, and is turned into traditional analytics and AI, whether via SQL, Python, or AI agents.
Growth, founders, and achievements
Databricks was founded by the original creators of Apache Spark out of UC Berkeley, led by CEO Ali Ghodsi. The company has scaled into one of the fastest‑growing large software firms in the world:
Revenue run‑rate above 5.4B dollars, growing more than 65 percent year‑over‑year as of early 2026.
AI products alone already at roughly 1.4B dollars run‑rate, with very high net revenue retention above 140 percent, signaling strong expansion within existing customers.
Recent financing at roughly 134B dollar valuation, positioning the company as a likely mega‑cap at IPO.
TAM, moat, and problems solved
Databricks attacks the massive market for data platforms, warehousing, and enterprise AI – the same multi‑hundred‑billion spend buckets that used to belong to legacy warehouse vendors, big data platforms, and ETL/tooling sprawl.
Its moat is built on:
Unified Lakehouse: the ability to store raw and structured data in one architecture with ACID transactions and strong performance, reducing the need for separate lakes and warehouses.
Unity Catalog: centralized governance, lineage, and permissions across all data and AI assets, which is crucial for large enterprises and regulated industries.
Open, multi‑cloud AI integration: native support for leading models, including Anthropic’s Claude model family, exposed directly via SQL or endpoints inside the platform.
The problem Databricks solves is data and AI fragmentation. In most enterprises, data lives in lakes, warehouses, marts, and spreadsheets, with separate tools for ML, analytics, and governance. Databricks consolidates this into a single Data Intelligence Platform where pipelines, features, models, and AI agents share the same governed foundation.
Why Databricks is a top private pick
For G.L.A., Databricks is one of the top three favorite private companies, alongside other AI infrastructure and defense‑tech plays. With a huge run‑rate, strong AI monetization, and a likely IPO in the not‑too‑distant future, it represents one of the cleanest ways to own the “picks and shovels” of data and AI without betting on any single model or application. The platform is disrupting both old‑school warehouses and point‑solution ML tools, not the other way around.
Why Their Partnership Matters More Than Their Competition
Investors often ask “Do Palantir and Databricks compete?” The more useful framing: they are complementary layers of a modern data and AI stack, and their strategic product partnership formalizes what joint customers were already forcing into existence.
Core pillars of the integration
The partnership, announced in March 2025 and deepened since, centers on four pillars of product integration: data federation, governance integration, compute pushdown, and AI and workflow integration. In practice, that means:
Secure, bidirectional data federation: Foundry can register Databricks tables as “virtual tables” via Unity Catalog, and Databricks can register Palantir’s catalogs, allowing each side to access data without copying it.
Unified governance: Unity Catalog provides the unified data governance and credential vending, while Palantir overlays its multimodal security and ontology‑level permissions.
Compute where the data lives: pipelines authored in Palantir can run their compute in Databricks’ engine (compute pushdown), lowering latency and TCO while keeping data in place.
AI and workflows: models trained in Databricks can be registered and invoked from Palantir AIP workflows, combining Databricks’ training and feature tooling with Palantir’s operational AI applications.
For customers, this means they can:
Keep a single copy of data in the Databricks Lakehouse.
Use Palantir to model their enterprise, build operational apps, and orchestrate AI workflows.
Use Databricks for heavy compute, ETL, warehousing, and LLM fine‑tuning.
Data is the new gold, and in the AI era, companies with the largest and highest quality its of proprietary data will be the big winners. “Companies that will prevail in the AI era are those with AI expertise and pools of proprietary data” – Cathie Wood
Why They Are Disrupting, Not Being Disrupted
Both companies are structurally aligned with where the world is going: more data, stricter governance, and AI embedded directly into operations rather than bolted on.
Palantir is disrupting the “PowerPoint and consultants” model of digital transformation and the fragmented BI/ETL/tool soup by offering an operating system for AI‑driven decisions that runs on top of customers’ existing data estates.
Databricks is disrupting “lake + warehouse + ML add‑ons” by collapsing them into a governed Lakehouse and Data Intelligence Platform that can run everything from batch ETL to AI agents, with economics and performance tuned for cloud‑scale workloads.
Together, their integration gives large enterprises a blueprint: Databricks as the data and AI engine, Palantir as the operational AI and decision layer, joined by secure, bidirectional data and compute flows.
“Bottom Line, if you missed $PLTR’s last run, this may be a time to think about entering before their next earnings , more importantly this may be the time to buy and hold $PLTR as becomes a trillion dollar company in the not so distant future” – Coach Kyle
