⚙ Agent Engineering · for business

When the promptstopsworking

I design and build production agent systems: orchestration, observability, fallback routing, recovery. Not "one big prompt" — engineering architecture with verification and traces.

→ Discovery call ⬡ Open course

STATUS: MONITORING AGENTIC_FLOW

NODE_01: OK

NODE_02: OK

ORCHESTRATOR: OK

RECOVERY: OK

// services

Three engagement formats

RATIONALE: Мы не строим промпты. Мы инсталлируем автономную инженерную инфраструктуру.

SVC-SKILL-MOD-01

Audit & Spec

Map your manual processes → architectural spec for an agent system. Includes AI/Tool/Code node decomposition and ROI assessment.

Output

Implementation-ready spec

Term

~2 weeks

SVC-SKILL-MOD-02

Build & Ship

Pipeline implementation on n8n / Claude Code / custom backend. CI/CD, observability via Langfuse, LLM fallback routes, alerting.

Output

Production-ready system

Term

~6 weeks

SVC-SKILL-MOD-03

Embed Engineer

Part-time embedded agent engineer with your team. Help with design, reviews, fighting production fires.

Output

Continuous capacity

Term

monthly retainer

// cases

What I’m already building

ENERV

personal knowledge pipeline

A semantic journal across 1400+ Markdown nodes. Gmail Trigger → Gemini Router → ACL Transformer → Obsidian vault. 9 nodes, 2 LLMs, idempotent via Gmail labels.

stack: n8n · Gemini · Qdrant · Docker

SOVERN

production infrastructure

Self-hosted agent stack: Hetzner CX22 + Cloudflare Tunnel + Docker bridge. n8n + Langfuse + Postgres. LiteLLM 5-way fallback routing, SMTP via Resend.

stack: Hetzner · CF Tunnel · Docker · LiteLLM

// process

How we work

STEP_01

Discovery call (30 min, free)

STEP_02

Workshop: task decomposition, scope, acceptance criteria

STEP_03

Spec → your team or I implement

STEP_04

Build sprints with weekly check-ins

STEP_05

Handover + observability + runbook

Let’s start with a conversation

30 minutes to understand the task and whether it makes sense to work together. No fluff, no slides.

mamaev.sasha@gmail.com