PST.AG
Website:
pst.ag
Job details:
We're building the next generation of autonomous Hermes Agent AI systems — agentic architectures powered by the Model Context Protocol (MCP). We're looking for a Technical Lead who can architect, ship, and own agentic AI solutions end-to-end. This is a hands-on technical leadership role first — you'll be in the code, shaping agent behavior, designing tool-use flows, and making real-time decisions about safety, cost, and reliability.
JOB LINK: https://erp.pst.ag/jobs/pst.ag/it-project-manager-ai
What You'll Do
- Architect & build agentic AI systems — ReAct loops, planner-executor patterns, multi-agent orchestration
- 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, and unblocking delivery
- Drive agent safety & reliability — implement guardrails, handle non-determinism, manage token budgets, and monitor for goal misalignment
- Own production outcomes — from prototype to pilot to production, you're accountable for the system working in the real world
- Nice-to-Have (not required)
- Classical project management experience (WBS, Gantt, EVM, change control)
- PMP or PRINCE2 certification
- Formal waterfall or hybrid delivery background
What We're Looking For✅ Proven agentic AI delivery (Hermes Agent) — you've shipped at least one agentic AI system to production or pilot (autonomous support agent, multi-agent workflow, tool-using LLM, etc.)
✅ Deep understanding of agent architectures — ReAct, planner-executor, tool-use loops, memory, and orchestration
✅ Hands-on experience with MCP (or the ability to ramp quickly)
✅ Strong technical communication — you can explain architecture decisions, trade-offs, and failure modes to both 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
Why Join Us?
- Work on the cutting edge of agentic AI and MCP — today's frontier, not yesterday's enterprise
- High ownership, low bureaucracy
- Competitive compensation
Click on Apply to know more.