Sekou M. Doumbouya - Technical Insights

I ran official Qwen3.8 27B BF16 and FP8 side by side, then changed one serving knob at a time.

The best long-context profile I measured passed every synthetic gate and crashed twice under real traffic.

Why I added a second RTX PRO 6000, upgraded my CPU, and then watched DeepSeek 0731 land on the exact card I just bought.

A 41-task graph ran for 41 hours, produced 115 reviews, and merged nothing.

How wiring two mismatched GPUs into my coding harnesses turned into intent-based routing, and why the measurement, not the transport, became the product.

anvil-serving routes coding-agent work to local models only where measured quality has earned it, verifies local output, and escalates up the tier ladder the moment verification fails.

Why I built a dual-GPU workstation and stood up local model serving, and why owning the durable layer is the bet I am most glad I made.

Two days of autonomous agents shipped a release and moved a billion tokens.

An agent reported a test it never wrote.

I handed an agent a goal and one sentence, then slept. 23.7 hours of unattended runtime later, Anvil v0.3.0 was on PyPI.

The question that routes an agent task is not which model is best.

A single 5090 could run three local agents.

I ported a nine-agent crew from one harness to another.

Maxing my AI plans and fighting rate limits, I split the spend across providers, bought a local GPU for overflow, and designed a pull queue for my agents.

Every spec-driven tool stops at the same place: a great document handed to an agent, and trust.

Spec-driven development feels like a 2025 invention.

Spec Kit, Kiro, Tessl, task-master, BMAD, Agent OS, OpenSpec: seven projects, twelve months, one idea.

I mined four hundred GitHub issues from the tools next to mine.

I had an agent system audit my agent system against the field.

Three plugins, four invariants, one principle.

A local-first project-state layer in SQLite, with claims that have heartbeats and evidence captured by hooks.

Workflow orchestration alone won't save you.

Agent SDKs give you tools and reasoning.

A 10-package monorepo, 27 MCP tools, checkpoint-based durability, and three sandbox types.

How I went from managing AI agent context windows by hand to building an orchestration system that lets me describe what I want and walk away.

What I learned building specialized agent archetypes that coordinate through file ownership, wave execution, and structured handoffs.

After twenty years building cloud platforms, here's what I've learned about bridging into AI infrastructure, and which skills matter more than you'd expect.

How a structured competitive analysis of my own code convinced me that a clean-room rewrite was the right call, and what I learned building Nexus.

Why the blueprint document I wrote for our AWS organization might be the highest-leverage thing I shipped that year.

The only sustainable scaling strategy is multiplying competent decision-makers.

Why treating your internal platform like a product changes how infrastructure teams operate, and how golden-path solutions reduce cognitive load.

Why periodically questioning inherited infrastructure decisions matters, illustrated through an AWS Transit Gateway to VPC peering migration.

Why staying hands-on at the principal engineer level creates better outcomes for your team, your architecture, and your credibility.

A taxonomy of region types, spec-driven provisioning, and phased orchestration for predictable multi-region expansion.