• Sourabh AI
  • Posts
  • AI’s New Power Play: $30B Cloud Deals, $100M Search Wars, and Agent Startups Scaling Fast

AI’s New Power Play: $30B Cloud Deals, $100M Search Wars, and Agent Startups Scaling Fast

OpenAI teams up with Oracle, Perplexity eyes India’s 360M users, Google bets on hydropower for AI, and agent-based compliance startups hit $300M valuations; discover what’s next in the AI race.

Hey AI Techies,

AI development is accelerating at the intersection of compute, capital, and compliance. This week’s updates reveal how infrastructure investments and agent-powered models are redefining the next phase of AI.

Let’s explore the key developments shaping the landscape.

1. OpenAI and Oracle Strike $30 Billion Deal for AI Compute Expansion

OpenAI has signed a long-term agreement with Oracle, reportedly worth $30 billion, to secure additional computing power for training and deploying advanced AI models.

  • The partnership gives OpenAI access to Oracle Cloud Infrastructure (OCI) and expands its footprint beyond Microsoft Azure.

  • Oracle’s Gen2 architecture supports Nvidia’s H100 GPUs, optimized for high-throughput AI workloads.

  • This move strengthens OpenAI’s independence in compute sourcing while reducing reliance on a single cloud vendor.

Insight: The deal signals the growing demand for multi-cloud infrastructure in supporting foundation model training. Compute access is becoming a core strategic asset for AI companies.

2. Perplexity Raises $100 Million, Targets India’s 360 Million Internet Users

Perplexity, the AI-native search engine, has raised $100 million in fresh funding, with plans to scale in India.

  • The company is positioning itself as a credible alternative to Google by combining retrieval-augmented generation (RAG) with conversational interfaces.

  • India is a key market, with over 360 million active internet users seeking faster, context-aware search.

  • The launch of localized language models is expected in Q4 2025 to support vernacular queries.

Insight: The search experience is being reimagined around LLMs. Perplexity’s India-first strategy shows how region-specific AI applications are unlocking new growth.

3. Google Commits $3 Billion to Hydropower Infrastructure for AI Scaling

To meet its growing AI compute needs, Google has committed $3 billion to fund hydropower projects in the US and Europe.

  • The investment will support clean energy sourcing for Google’s AI data centers.

  • This is part of Alphabet’s broader push to meet its 2030 goal of running on 24/7 carbon-free energy.

  • Hydropower is being prioritized due to its reliability and compatibility with AI’s high-power requirements.

Insight: As AI models grow more resource-intensive, energy sourcing will become a differentiator. This move reflects a long-term strategy to align AI growth with environmental responsibility.

4. MIT Dropouts Build $300M AI Compliance Startup Using Agentic Workflows

Two MIT dropouts have built a $300 million valuation startup that automates compliance processes using AI agents.

  • The platform uses fine-tuned LLMs to handle legal documentation, risk scoring, and audit preparation.

  • Enterprises are adopting agent-powered compliance to reduce operational costs and improve regulatory accuracy.

  • Their product integrates with major enterprise systems like Salesforce and SAP.

Insight: Agentic AI is no longer experimental. It’s being deployed in highly sensitive industries like finance and legal, indicating maturity and trust in the agent-based automation model.

AI Tool of the Week

  • Tool: AgentOps
    Use Case: Manage and monitor production-grade AI agents

    • Features: agent analytics, fallback logic, workflow orchestration

    • Trusted by early adopters in legaltech and fintech

    • Free tier available for developers

Prompt of the Week

Boost Compliance with AI Agents
Prompt:
"Act as an enterprise compliance officer. Build an internal AI agent that automatically classifies documents based on regulatory risk, assigns escalation levels, and suggests mitigation workflows. Output a step-by-step implementation strategy using open-source LLMs."

✍️ About the Author

Sourabh Joshi is an AI thought leader, educator, and writer with over 18,000+ followers on LinkedIn and 6000+ subscribers to the Sourabh AI newsletter. He helps tech professionals and founders stay ahead with insights on Generative AI, solopreneurship, and the future of work.

📬 Follow on LinkedIn: linkedin.com/in/iam-sourabh-joshi