As of July 2, 2026, investor attention in the technology sector is pivoting from GPU-centric hardware scaling toward the "Agentic AI" era. Ivana Delevska, founder and CIO of SPEAR Invest, identifies this phase as a transition toward inference and autonomous software agents, requiring a fundamental reallocation of capital across the broader technology stack.
| Targeted Sector | Investment Rationale |
|---|---|
| Connectivity | High-speed demands of agentic workloads. |
| Semiconductors | Rising wafer complexity and process intensity. |
| Security | Protecting distributed autonomous environments. |
Core Stock Picks and Strategic Positioning
In her July 2 appearance on BNN Bloomberg’s Market Call, Delevska outlined three primary equities positioned to capture value from this architectural shift:
Astera Labs: Identified for its capacity to handle high-speed data movement, moving beyond traditional GPU dependency to satisfy increased inference bandwidth requirements.
KLA Corp: Selected as a play on the rising difficulty of chip manufacturing; as infrastructure demand climbs, the need for advanced process control and diagnostic precision intensifies.
Cloudflare: Positioned to capture the requirement for low-latency, secure, and distributed computing environments necessary for agents to function outside of centralized data centers.
Structural Market Observations
The current investment thesis argues that "Agentic AI" generates a distinct surge in inference workloads, fundamentally different from the training-heavy phases that characterized previous years.
"As AI shifts toward inference and autonomous agents, investment is broadening across the technology stack, creating new opportunities in CPUs, memory, networking, and cybersecurity." — Ivana Delevska
This assessment aligns with a broader, long-term outlook held by Delevska regarding the Data Center value chain. Historical records of her analysis show a persistent focus on identifying points of friction within the supply chain, moving from general cloud spending assessments in 2023 toward specific hardware bottlenecks today.
Read More: VALR integrates decentralized tech for new trading options from July 6th
Contextual Trajectory
The movement of capital into inference-based infrastructure follows a series of market cycles identified by Delevska since at least late 2024. Her previous calls—often emphasizing that AI stocks are not homogenous—have consistently warned against viewing the sector as a singular block. The current shift to "Agentic" systems marks an evolution from simple generative outputs to automated workflows, which necessitates a more granular look at memory bandwidth and secure network architecture than the initial "build-out" phase of 2024–2025.
Disclaimer: This summary reflects recent market commentary provided on July 2, 2026; it is not an endorsement of the assets mentioned.
