Private Market Weekly” – 4 Private Companies Im Investing In

“Private Market Weekly”

4 Private Companies I’m Investing In

May 2026 Edition | by Coach Kyle and G.L.A INVE$T

“The greatest wealth-building opportunities of the next decade will not be found on a public exchange. They will be found here — in the private markets — where the infrastructure of the AI economy is being built by a small number of exceptional companies, before the world catches on.”

The Fab 4 in Todays Spotlight

Coach Kyle’s Note

What these four companies share is not a sector — it is a moment. We are living through the early stages of the greatest innovation boom since the advent of electricity and the automobile. The full replatforming of the global economy around artificial intelligence, and each of these companies is building a foundational layer of that new infrastructure: the data and AI backbone (Databricks), the search and agent interface layer (Perplexity), the medical intelligence system (Open Evidence), and the autonomous physical hardware (Saronic).

This is not the internet boom and bust where companies were valued by potential alone, These are revenue-generating, commercially deployed, category-defining businesses — each with a credible path to becoming one of the most valuable companies in its respective sector within the next 5–10 years. One of the G.L.A Invest core principles is , “we only invest in companies our research states has the potential to 10X , 20X, or 30X over the next 5-10 years, all 4 of these companies fits that mold perfectly”- Coach Kyle

The private market is where fortunes are made before the public finds out. This is that moment.

01 — DATABRICKS

The Data & AI Infrastructure King

Snapshot

Why I’m Invested

Databricks is the picks-and-shovels play for the entire AI economy. Every enterprise that wants to use AI needs to store, process, govern, and operationalize data first. Databricks is the company that does that — and it does it better than any competitor on the market.

The numbers are extraordinary. A $5.4 billion revenue run-rate growing at 65% year-over-year is a profile that most public software companies would dream about. The company is cash flow positive. It has 20,000 global enterprise customers — representing the Fortune 2000 and beyond — that are deeply embedded in its platform. These are not trial users. The average $1M+ customer (650 of them) has an entire data and AI infrastructure built on Databricks. That is not a customer you switch away from easily.

The valuation step-up tells its own story: $62 billion at the start of 2025 → $100 billion (Series K, August 2025) → $134 billion (Series L, December 2025). In one calendar year, Databricks more than doubled in valuation while growing revenue 65% and turning cash flow positive. That is an exceptional combination.

The Product Suite

Databricks Data Intelligence Platform
The unified lakehouse — combining data warehousing, data engineering, machine learning, and AI governance in one platform. The architecture eliminates the need for separate data warehouse + data lake + ML infrastructure, which is why enterprises have consolidated on it.

Mosaic AI
The enterprise AI development layer. Includes fine-tuning, model serving, vector search, and AI agent development tools. Mosaic AI crossed $1 billion in revenue run-rate — Databricks now has two separate $1B+ product lines.

Lakebase (newest product)
A serverless database built for AI agents — launched with the $1 billion acquisition of Neon (January 2026). Growing at twice the pace of the data warehousing business in its first six months. As AI agents require real-time transactional databases, Lakebase positions Databricks as the default data layer for the agentic AI era.

Genie
The natural language analytics interface — ask questions about your data in plain English, get answers as tables, charts, or written summaries. Genie monthly active users grew >300% year-over-year on Azure Databricks alone. 98% of Databricks SQL warehouse customers now use the AI/BI features.

Delta Lake / Delta Sharing (Open-Source)
The open-source foundation under the entire platform — used across the industry, creating a massive developer ecosystem and standardization moat that drives commercial Databricks adoption.

The IPO

Ali Ghodsi has stated a 2026 IPO is possible. No S-1 has been filed as of May 2026, but Nasdaq is the expected venue. At $134B valuation with $5.4B ARR, Databricks would be one of the largest software IPOs in history. For pre-IPO investors, the current entry point is the last available before public pricing — which, for a company with this financial profile, could open significantly above the last tender offer price.

Key Risks

• Valuation at $134B requires sustained execution; any revenue deceleration will be punished

• Competition from Snowflake, Microsoft Fabric, Google BigQuery on specific workloads

• IPO timing uncertainty — private investors face illiquidity until listing

02 — PERPLEXITY AI

The Answer Engine That Became an Agent Platform

Snapshot

Why I’m Invested

Perplexity started as an AI search engine — the challenger to Google. It has evolved into something larger: a full-stack AI agent company that earns revenue from search, subscriptions, enterprise tools, and now agentic computing. The pivot is working. ARR went from ~$100M in early 2025 to $500M by April 2026 — a 335% year-over-year increase — driven primarily by the launch of Perplexity Computer, its autonomous agent product, in February 2026.

This is not a one-trick company. Perplexity has built a multi-layer AI consumption platform: the search layer (where the brand was built), the subscription layer (Perplexity Pro), the enterprise layer (API and business plans), and now the agent layer (Perplexity Computer). Each layer is additive to revenue and user engagement.

The Microsoft partnership — $750 million over 3 years — provides not just capital but access to Azure compute at scale and distribution inside Microsoft’s ecosystem, giving Perplexity the infrastructure to compete at a level that its $1.5B in total funding alone could not support.

The Valuation Arc

The trajectory from $500M to $20B in two years is exceptional. What is notable is that the valuation is not running ahead of the business — revenue is accelerating to meet it.

The Product That Changed Everything: Perplexity Computer

Launched February 2026. An autonomous AI agent that can operate computers — browsing the web, writing and executing code, filling forms, drafting documents — on the user’s behalf.( I use this product all day!) Think of it as the operating layer for AI agents, built on top of the search and knowledge infrastructure Perplexity already owns.

The revenue impact was immediate: ARR jumped from ~$200M at the end of 2025 to $450M in March 2026 and $500M in April 2026 — a 50% jump in a single month following the agent pivot. This is the kind of product-market fit signal that justifies the valuation premium.

The Comet Browser (Perplexity’s AI-native browser) extends the same thesis into the consumer’s daily computing environment — a distribution surface that could rival or complement Google Chrome at scale.

The Strategic Position

Perplexity’s edge is not just technology — it is data and habit. The company is training on search queries from millions of users who are, by definition, trying to learn things. That intent-driven dataset is one of the most commercially valuable collections of human curiosity in existence. Combined with the Microsoft Azure infrastructure deal and multi-model access (OpenAI, Anthropic, DeepSeek, xAI via the Foundry platform), Perplexity has the AI stack depth of a company 10x its size.

Key Risks

• Google, OpenAI, and Microsoft are all building competing AI search and agent products with greater distribution

• No confirmed IPO timeline; $20B valuation demands continued growth execution

• Heavy compute costs; profitability path still being built

03 — OPEN EVIDENCE

The AI Brain for American Medicine

Snapshot

Why I’m Invested

Open Evidence is the single most impressive adoption story in AI applied to a professional domain. In the history of healthcare technology, no platform has ever reached 40% of all U.S. physicians within its first few years of operation. That is not a marketing claim — it is documented in peer-reviewed literature and confirmed by Reuters and Sutter Health partnership data.

The founder has done this before. Daniel Nadler built Kensho — the AI financial intelligence platform — and sold it to S&P Global for $550 million. Then he recognized that medicine had the same information overload problem that financial markets did, and built the same solution for doctors. Open Evidence is Kensho for medicine. The parallel is almost exact, and it is a thesis that the best venture capitalists in the world have validated: Thrive, Sequoia, Google Ventures, Kleiner Perkins, DST, Blackstone, NVIDIA, and the Mayo Clinic itself are all investors.

The Valuation Ladder — One of the Fastest in History

12x valuation growth in under one year. That is not a fluke — it reflects the most important metric in healthcare AI: real physician adoption that cannot be reverse-engineered by a competitor once it becomes habitual and embedded in clinical workflow.

The Product

Medical Search (Core)
Physicians ask clinical questions in natural language. Open Evidence searches 35 million peer-reviewed publications and returns cited, evidence-based answers. Critically — if the evidence is inconclusive, the platform will not answer, eliminating the hallucination risk that makes general AI dangerous in clinical settings.

Open Evidence is the only AI platform to achieve a perfect score on the United States Medical Licensing Examination. Content partnerships with the New England Journal of Medicine and the Journal of the American Medical Association (11 specialty journals) mean Open Evidence has access to the most authoritative medical literature in the world — with exclusive content that competitors cannot replicate.

OpenEvidence 2.0
Adds clinical calculators (50+), patient handout generation, and prior authorization letter drafting. A physician who previously needed 5 separate apps now needs one.

OpenEvidence Visits
Real-time patient encounter support. As the doctor speaks with a patient, OpenEvidence enriches clinical notes, surfaces relevant treatment guidelines, and embeds evidence into the documentation workflow. This is the product that turns Open Evidence from a search tool into a clinical operating system.

The Business Model

Open Evidence is free for all verified U.S. physicians. Revenue is generated through pharmaceutical advertising — contextually placed at the point of clinical decision-making. A doctor asking about diabetes treatments may see a clearly labeled sponsored module from a relevant pharmaceutical manufacturer. This is the Google AdWords model applied to healthcare at scale, at the highest-value moment of physician attention.

The U.S. pharmaceutical industry spends approximately $30 billion annually marketing to healthcare professionals. Open Evidence — reaching 40% of U.S. physicians at the moment they are actively making clinical decisions — is the highest-value digital channel that market has ever had access to.

Future revenue streams: CME credits, EHR integration, international expansion, and potential AI licensing to hospital systems.

Key Risks

• Advertising revenue model requires trust — any perception that ads influence clinical answers would be catastrophic

• UpToDate (Wolters Kluwer) and Epic Systems have institutional relationships and procurement leverage that Open Evidence still lacks

• HIPAA compliance and data governance must be maintained at scale as platform deepens into patient encounter workflows

• International expansion requires clinical localization for each market’s guidelines and regulations

04 — SARONIC TECHNOLOGIES

The Autonomous Navy America Never Had

Snapshot

Why I’m Invested

Saronic is building something that has never existed before at scale: a fleet of autonomous surface vessels, manufactured domestically, operated by AI, and contracted directly by the U.S. Navy. The company has done in four years what traditional defense primes take decades to accomplish. It has moved from prototype to large-scale Navy production, built a purpose-designed shipyard in Louisiana, assembled a world-class team, and established itself as the default U.S. partner for autonomous surface vessels.

The $392 million Navy contract through 2031 is not a research grant or a study. It is a production contract. Saronic is delivering vessels. That is a moat that no competitor can replicate quickly — institutional trust with NAVSEA is earned through performance, not capital.

The data flywheel is the core long-term asset. Every Corsair at sea generates training data that improves the autonomy stack. With 100,000+ miles logged, Saronic’s models are trained on more real-world maritime operational data than any competitor. This gap widens with every hour of operation.

The Product Architecture

Corsair (24 ft) — The high-rate production platform. The workhorse of the Navy contract. Already in mass production with 100,000+ operational miles logged.

Marauder (180 ft) — The autonomous ship. Functionally comparable to an offshore support vessel, but with no crew requirements — meaning 30% more fuel capacity, lower operating costs, and the ability to remain at sea indefinitely. Four hulls in simultaneous production.

Future Pipeline — Panamax-class ambitions (600–1,100+ ft). Autonomy applied to large commercial shipping would be one of the most disruptive events in the history of global trade.

The Defense Thesis

The U.S. Navy is moving toward a “hybrid fleet” doctrine — manned warships extended and protected by swarms of autonomous surface vessels. The Strait of Hormuz, South China Sea, and Arctic routes are all active theaters where ASVs extend force projection while keeping personnel out of harm’s way.

Saronic has embedded relationships at NAVSEA, the Defense Innovation Unit, and U.S. Special Operations Command. That institutional access — earned through performance — is the single hardest thing to replicate in defense technology.

Key Risks

• Regulatory frameworks for autonomous vessels are still maturing (USCG, IMO, ABS)

• Commercial maritime adoption moves slower than defense due to operator trust curve

• Scaling production requires skilled maritime trades labor that is currently constrained

• At $9.25B, valuation assumes continued flawless Navy contract execution

Portfolio Thesis: The Bigger Picture

These four investments, taken together, represent a concentrated bet on the infrastructure layer of the AI economy — covering data (Databricks), information and agents (Perplexity), medical intelligence (Open Evidence), and autonomous physical systems (Saronic).

Each one is the best-positioned company in its respective category. Each one has validated technology, real revenue, and elite institutional backing. Each one is still private — meaning the largest gains are still ahead of public market discovery.

The common thread: we are not investing in AI as a concept. We are investing in companies that have already won their initial categories, are generating real revenue, and are now expanding into the broader AI economy that will define the next 20 years.

Until Next Time, Coach Kyle out!

G.L.A. Invest Research — For informational purposes only. Not financial advice. Past performance does not guarantee future results. Private market investments carry illiquidity risk and are suitable only for qualified investors.