Technical Lead

Technical Lead – Agentic AI & MCP 🤖⚡

📍 Remote / Full-Time

🧑 💻 PST.AG

Job link: https://erp.pst.ag/jobs/pst.ag/it-project-manager-ai

About the Role

We're building production-grade autonomous AI systems on the Hermes Agent platform by Nous Research. We need a Technical Lead who has shipped with Hermes — not just read about it. You'll architect agent workflows, wire up MCP tool servers, and own the delivery of multi-agent systems from prototype to production.

Must-Have — Hermes Agent Experience

✅ You have shipped a production or pilot project on Hermes Agent — skill files, cron jobs, task delegation, Kanban auto-dispatch, multi-agent orchestration

✅ You know how to build and manage Hermes subagents — parallel task delegation, failure retry, context injection

✅ You've wired up MCP connectors to tools like Slack, Outlook, Jira, ADO, or custom APIs within the Hermes framework

✅ You understand the Hermes architecture — agent lifecycle, provider-agnostic design, session memory, workflow learning

What You'll Do

Architect & build agentic AI systems on the Hermes platform — ReAct loops, planner-executor, multi-agent workflows

Design & integrate MCP tool servers and registries for real-world tool-using LLM applications

Lead a team of ML engineers, MCP developers, and safety testers — setting technical direction, reviewing architecture, unblocking delivery

Drive agent safety & reliability — guardrails, non-determinism handling, token budget management, goal misalignment detection

Own production outcomes — from prototype to pilot to production, you're accountable for the system working in the real world

Build reusable workflows — capture patterns as Hermes skills so the team ships faster over time

Nice-to-Have (not required)Formal project management experience (WBS, Gantt, EVM, change control)

PMP or PRINCE2 certification

Traditional waterfall/hybrid delivery background

What Else We're Looking For

✅ Deep understanding of agent architectures (ReAct, planner-executor, tool-use loops, memory, orchestration)

✅ Strong technical communication — explain architecture decisions, trade-offs, and failure modes to engineers and stakeholders

✅ Structured, accountable, proactive — you own outcomes and lead by example

Bonus Points

Built safety guardrails for LLM agents (prompt injection defenses, rate limiting, tool-use policies)

Managed agent cost optimization (token tracking, model routing, caching strategies)

Experience with multi-agent orchestration frameworks beyond Hermes

Why Join Us?

Work on the cutting edge of Hermes Agent + MCP — you'll be shaping the platform, not just using it

High ownership, low bureaucracy