Tony Bai's tech world (tonybai.com). Not satisfied with just "knowing how", we strive for mastery. Focused on Go language internals, high-quality engineering practices, and cloud‑native architecture, exploring cutting‑edge intersections of Go and AI. Gophers who pursue technology are welcome—follow me and evolve with Go.
OpenAI's Forward Deployed Engineers (FDEs) bridge the gap between AI demos and production by embedding with customers, turning custom solutions into reusable capabilities, and feeding real-world model gaps back to research—aiming to make themselves unnecessary as models improve.
This article dissects VictoriaLogs' performance engineering across four layers—data architecture (LSM-tree, time partitioning, log-stream locality), query optimization (sparse indexes, Bloom filters), columnar storage with type-specific encoding and compression, and Go runtime tuning—showing that reducing unnecessary work via smart data organization outweighs micro-optimizations.
This article explores how formal verification tools like Gobra and Viper enable machine-checked correctness proofs for Go programs, using preconditions, postconditions, loop invariants, and permission logic to verify functional correctness, memory safety, and concurrency properties, and discusses how AI agents can assist in generating specifications while verification tools provide independent guarantees.
This article uses a chef-kitchen-restaurant analogy to explain the Harness concept in AI Agent systems, distinguishing Interface, Harness, MCP, and Agent roles, and detailing how LLMs use tools via Harness with examples like Claude Code, Codex, and T3 Code.
This article explores Go's new experimental SIMD packages (simd/archsimd and simd), explaining SIMD hardware concepts, vector register operations, common SIMD patterns like broadcast and fused multiply-add, compiler intrinsic mapping, and demonstrates a dot product implementation achieving 9x speedup over scalar code.
A backend engineer abandons Svelte for Go + HTMX + templ + sqlc, eliminating frontend state sync bugs, enabling type-safe HTML partials, and improving LLM code reviewability, with practical patterns for search, lazy loading, OOB swaps, and auth redirects.
Go proposal #77441 adds Server.HandleRequest to net/http, exposing internal request-processing middleware (OPTIONS *, DefaultServeMux fallback, panic recovery) so HTTP/3 and other protocol implementations can reuse standard library behavior without code duplication, while also enabling full-server integration testing.
Google's Harness² introduces a dual-layer recursive self-improvement framework where agents optimize their own harnesses—prompt, tools, memory, skills—via behavioral comparison instead of ground truth, achieving up to 11.5-point gains on benchmarks and enabling weaker models to surpass stronger ones with static harnesses.
Go security team reports monthly vulnerability submissions jumped from 5–10 to 100–200, mostly false positives from AI tools; they respond with re-prioritization, LLM-assisted triage, and shift-left CI checks while refusing to adopt Linux's full-disclosure stance.
The Claude Code team at Anthropic shares their evolved code review process where AI agents write 90% of code, using automated gates, risk-based human approvals, merge queues, feature flags, and incident-driven evals to cut revert rates by 40% while scaling PR volume 10x.
