In theory, there is no difference between theory and practice. In practice, there is. - Yogi Berra
Building data-intensive, real-time systems.
Data-intensive applications on a streaming backbone—not a batch afterthought.
Kappa-shaped foundations and open table formats—one processing model.
Sub-second insight and inference without overnight copy-and-relayout.
CDC, event mesh, and always-on sync across systems of record and engagement.
AI: From peak hype to pragmatic productivity.
Durable multi-agent plans—because MCP tool calls fail.
In-harness assistants, eval loops, and guardrails that keep agents steerable.
Domain-specific models—RL’d, fine-tuned, distilled—for cost, control, and tokens.
Good tools for in-harness use—not another self-branded agent by default.
Operating like a digital native. A commitment to self-service everything.
Traffic plane for cross-cutting—N–S and E–W, APIs, mesh, MCPs, agentic sprawl.
Standardized telemetry at agent/botnet scale—without breaking the bank.
Guardrails for token-maxxing teams—agents, domains, runtimes, humans.
When incremental optimization will not cut it—a techno-business worldview rooted in numbers.
Zero-downtime, active-active data and application moves.
Tail optimization for distributed and concurrent systems.
First-principles fin-ops—architecture-shaped economics.
Distributed systems built for the problem, not the catalog.
