AI Agent Architecture Review
Budget / Salary$30–250
TypeFreelance project
LocationRemote
Posted3 hours ago
[Hiring] Architect who can read an AI agents codebase and explain it to engineers and non-engineers
**What it is**
AgentS is a multi-agent platform in beta. A user asks for something by chat or voice, an orchestrator turns the request into a graph of steps, and each step runs a versioned operation (research, analysis, writing, video). Some steps stop at human approval gates before they continue. Each customer's data is isolated from the others.
Stack: TypeScript monorepo. SvelteKit front end, Hono API on Bun, Postgres with pgvector and row-level security, Inngest for durable execution, LLMs through an AI gateway, Deepgram for voice, and an MCP-based tool layer.
**What I need**
Someone who reads the code, understands how it really works and explains it clearly. Deliverables:
1. An architecture map: components, data flow and the main path (request → graph → steps → human gate → deliverable).
2. A written document with diagrams, in two versions: one for engineers and one for non-technical readers.
3. An honest assessment: what is solid, what is risky, what you would simplify.
4. A recorded walkthrough call.
**You**
- Have designed or reviewed production systems built on LLMs or agents.
- Know durable workflow engines (Inngest, Temporal or similar) and multi-tenant Postgres.
- Write and draw clearly. Spanish is a plus, not required.
**Details**
Remote, fixed scope. NDA and read-only access to the repository. Budget: [X USD]. Timeline: [N weeks].
**To apply**, share this:
- A system you architected or reviewed that used LLMs or agents.
- A document or diagram you made that explains a complex system.
- Your rate and availability.
Start your message with the word "agents" so I know you read this far.
**What it is**
AgentS is a multi-agent platform in beta. A user asks for something by chat or voice, an orchestrator turns the request into a graph of steps, and each step runs a versioned operation (research, analysis, writing, video). Some steps stop at human approval gates before they continue. Each customer's data is isolated from the others.
Stack: TypeScript monorepo. SvelteKit front end, Hono API on Bun, Postgres with pgvector and row-level security, Inngest for durable execution, LLMs through an AI gateway, Deepgram for voice, and an MCP-based tool layer.
**What I need**
Someone who reads the code, understands how it really works and explains it clearly. Deliverables:
1. An architecture map: components, data flow and the main path (request → graph → steps → human gate → deliverable).
2. A written document with diagrams, in two versions: one for engineers and one for non-technical readers.
3. An honest assessment: what is solid, what is risky, what you would simplify.
4. A recorded walkthrough call.
**You**
- Have designed or reviewed production systems built on LLMs or agents.
- Know durable workflow engines (Inngest, Temporal or similar) and multi-tenant Postgres.
- Write and draw clearly. Spanish is a plus, not required.
**Details**
Remote, fixed scope. NDA and read-only access to the repository. Budget: [X USD]. Timeline: [N weeks].
**To apply**, share this:
- A system you architected or reviewed that used LLMs or agents.
- A document or diagram you made that explains a complex system.
- Your rate and availability.
Start your message with the word "agents" so I know you read this far.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.