# Lead AI Engineer, Agentic Systems — Catalyst·Wayfare

> Embed deep with clients and own the agent system end-to-end. Architecture, the engine, the
> patterns we replicate across engagements. Reports to the founder.

- **Team:** Engineering
- **Type:** Full-time or Contract — technical lead seat
- **Term:** 8 months, extension likely
- **Location:** Remote-first, US Eastern overlap. Monthly U.S. domestic travel (major metro).
- **Apply:** talent@cwai.co, subject line "Lead AI Engineer - Agentic Systems"

Canonical page: https://www.catalystwayfare.ai/jd-lead

## About Catalyst·Wayfare

Catalyst·Wayfare is an AI transformation firm that builds production AI systems for
mid-market enterprises in regulated and technical domains. We do not stop at PowerPoint AI
strategy. We ship working systems in partnership with our clients — and, where appropriate,
their engineering teams.

The team's backgrounds span MIT, McKinsey, the White House, and some of the most respected
names in tech.

## Why this role: embed deep. Own the agent system.

You would embed inside the engineering organization of a leading firm in a
critical-infrastructure sector, where the binding constraint on revenue is technical
throughput the labor market cannot supply. **You own the agent system that lifts it.**

This is a lead engineering seat. **GitHub history, not Salesforce dashboards.** You write the
code; we manage the room. The work is technically interesting (multi-agent orchestration over
real domain simulators, not chatbot demos) and commercially serious (review gates tied
directly to billable throughput).

We are looking for someone who **codes like an IC, communicates like a PM, and navigates
clients like a founder.** If you have run a small thing of your own, that is a strong signal.
You own architecture, the orchestration engine, and the agent design patterns we replicate at
the next client.

## How we ship: we use what we sell.

Claude and OpenAI APIs in production. Open-source models (Llama, Mistral, more) when the data
or the math points there. Cursor and Claude Code in our IDEs, daily.

Vercel, Neon, Sentry are some of our deploy surfaces. Modern infra, no six-month wait for IT
to approve a tool you already use at home.

Evals are first-class artifacts, not an afterthought. Agents we trust live behind audit
trails. **You will not be the engineer fighting a CISO to install Cursor.**

## What you will do

- **Lead architecture and build** of a multi-agent orchestration platform spanning roughly
  seven capability agents: data ingestion, requirements retrieval, model construction,
  simulation orchestration, QA, report generation.
- **Define interface contracts** with the client's software team for integrations into the
  domain-specific simulation tools their business depends on.
- **Design and build the orchestration engine:** state management, error recovery, audit
  trails, monitoring dashboards.
- **Build on existing data foundations:** a production RAG system with established retrieval
  infrastructure and live interaction telemetry.
- **Run a tiger-team pilot loop** with the client's senior engineers. Instrument, measure time
  savings, capture failure modes, refine.
- Embed periodically on-site with the client (heavier for the first quarter, lighter after).
  Travel is to a single U.S. major metro, monthly on average.
- **Mentor a junior builder** and shape technical hiring as we grow.

## What you will bring

### Must-haves

- Six plus years building production software, with two plus years shipping LLM-based or
  agentic systems in production (not POCs that died).
- AI-pilled and full-stack. You reach for agents instinctively and have opinions about which
  ones. But you got fluent with code first — Cursor multiplies you; it did not teach you.
- Strong Python. Comfortable with at least one agent framework (LangGraph, Letta, custom
  orchestration) and the discipline to know when to roll your own.
- Hands-on experience with RAG architectures: vector databases, embedding models, document
  processing pipelines, retrieval evals.
- Strong intuitions for LLM evals and agent reliability. You have debugged a system that
  worked 70 percent of the time and gotten it to 95.
- Comfortable as the technical voice in client-facing rooms with non-engineers, executives,
  and domain experts, translating between them.
- Clear written communication and willingness to overlap with (not mirror) US Eastern time
  for core collaboration hours.

### Nice-to-haves

- Prior client-embedded or solutions-engineering experience at an AI lab, applied firm, or
  similar.
- Experience integrating LLMs with deterministic engineering tools, simulators, or
  specialized APIs.
- Cloud infrastructure (AWS, Azure, GCP) for deploying production AI applications.
- Domain exposure to regulated or technical verticals (financial services, healthcare, legal,
  industrial).

### We are flexible on

- Years of experience if the work history shows the right shape. A strong mid-level engineer
  with a sharp track record of shipping agentic systems will be considered.
- Geography. See logistics above.

## Reporting

Reports to the Catalyst·Wayfare founder.

## How to apply

Send to **talent@cwai.co** with "Lead AI Engineer - Agentic Systems" in the subject line:

1. A short cover letter. Who you are, what you have shipped, why this seat.
2. A short note, three to five paragraphs, describing a multi-agent or production LLM system
   you built. What worked, what broke, what you would do differently.
3. GitHub, code samples, or CV — optional, only if they sharpen the above.

[All open roles](https://www.catalystwayfare.ai/careers)

---

_Catalyst·Wayfare — https://www.catalystwayfare.ai/ · [Agent guide](https://www.catalystwayfare.ai/llms.txt)_