zaks.io
Products
Trace Flow
LiveSee what your AI is actually costing you.
Your AI spend is a black box. You're shipping features using GPT-5, Claude, Gemini — but you have no idea what each call actually costs or how long it takes.
Trace Flow is a streaming proxy that sits between your app and your LLM providers. Change one base URL and get full visibility — costs, latency, and token usage across every provider.
Zero added latency
Streaming proxy architecture means your requests are never slowed down.
Multi-provider support
OpenAI, Anthropic, Google, and more — all in one dashboard.
Drop-in proxy, no SDK
Change one base URL. No code changes, no vendor lock-in.
Cost & usage dashboard
Real-time visibility into spend, token usage, and latency per model.
About
Built by Isaac Suttell
10 years building mission-critical infrastructure at PlayStation Studios. Every first-party AAA game shipped with my code. I was the one on call at 3am when a studio was in crunch and the tools went down.
I left to test a thesis: one engineer, augmented by AI, can build and ship software that rivals what large teams produce.
Zaks.io is that experiment. Bootstrapped. San Diego. No outside funding. Just good software.
Work With Me
Available for Select Engagements
I help teams design and ship AI-powered systems. If you're building with LLMs, agents, or automation and need someone who's done it at scale, let's talk.
AI Agent Architecture & Orchestration
Multi-agent systems, sandboxed execution, MCP servers
LLM Integration & Cost Optimization
Observability, prompt pipelines, multi-provider routing
Automation & Workflow Systems
Scheduled tasks, event-driven pipelines, edge compute