# EIS Group — Lead AI System Architect

| Field | Value |
|---|---|
| **Date found** | 2026-05-26 |
| **Company** | EIS Group |
| **Role** | Lead AI System Architect |
| **Location** | Remote |
| **Salary** | Undisclosed |
| **Job URL** | https://to.indeed.com/aa8gh8xvw77l |
| **Status** | New |

---

## Company Research

| Field | Value |
|---|---|
| **Headquarters** | San Francisco, CA, USA |
| **Founded** | 2008 |
| **Employees** | ~500–1,000 |
| **LinkedIn** | linkedin.com/company/eis-group |
| **Website** | https://eisgroup.com |
| **Blog** | — |

- **Product:** Open, API-first core insurance platform (Policy, Billing, Claims) enabling insurers to operate like a tech company — fast, composable, cloud-native.
- **Customers:** P&C and L&H insurers globally across all lines of business.
- **Notable:** Founded 2008, bootstrapped/PE-backed InsurTech; thousands of public APIs; clients in North America, Europe, APAC; "Scaled Agile" engineering culture; global distributed team.

---

## Job Summary

**What they do:** EIS Group builds a cloud-native insurance core systems platform that lets insurers launch products and automate workflows through an API-rich, composable architecture.

**The role:** Lead AI System Architect owning the full design of EIS's agentic AI platform — multi-agent orchestration, MCP tool ecosystems, agent memory and evaluation — across Policy, Billing, and Claims product domains, fully remote.

**Core work:**
- Own architecture of multi-agent orchestration system: agent patterns (ReAct, graph-based), MCP tool registry, agent memory (short/long-term/semantic), planning, safety, and evaluation harnesses
- Drive MCP strategy — which capabilities EIS exposes and consumes as MCP servers, tool versioning and schemas
- Define levels of agent autonomy (assistive → semi-autonomous → autonomous) with human-in-the-loop checkpoints for regulated insurance workflows

**Stack:** Python · LangChain · LangGraph · MCP · Vector databases · AWS / Azure / GCP · Java/Spring (preferred) · TypeScript

**Work style:** Fully remote; globally distributed team; "Scaled Agile" environment.

---

## Score: 80%

| Dimension | Score | Justification |
|---|---|---|
| Agentic AI depth (25%) | 88% | Deepest agentic scope seen in search — owns multi-agent architecture, MCP strategy, agent memory, planning, safety at platform level |
| Tech fit (25%) | 73% | LangChain/LangGraph, MCP, vector DBs, Python all aligned; Java/Spring listed as "strong" requirement — gap for Luca |
| Remote fit (25%) | 100% | Explicitly fully remote; globally distributed team; equipment and mobile allowances |
| Company culture fit (15%) | 62% | InsurTech product company, Scaled Agile, multicultural, modern engineering practices; not a startup but more nimble than big enterprise |
| IC/leadership balance (10%) | 58% | IC architecture ownership at the core, but significant leadership: mentorship, exec communication, cross-team alignment — more mixed than pure IC |
| **Final (weighted)** | **80%** | |

---

## Strengths

- Agentic AI depth is exceptional — this role architects the multi-agent platform end-to-end, not just implements features within it
- Fully remote with no ambiguity; equipment and mobile stipend provided
- LangGraph/LangChain and MCP stack alignment is strong; Luca's Fenergo agentic document pipeline maps directly to insurance automation workflows
- Insurance/compliance domain overlaps with Luca's KYC regulatory work at Fenergo

---

## Weaknesses & Risks

- Java/Spring listed as a "strong" skill alongside Python — Luca has no Java background; this is a real gap for a Java-first product company
- "Lead" title carries significant architectural communication to executives and cross-team mentorship duties — heavier leadership load than typical IC
- Salary undisclosed; InsurTech scale-up may offer below €110k target depending on location-based compensation bands
- Insurance domain has heavy compliance/audit requirements that could slow velocity (though the JD frames safety as a first-class concern, not a blocker)

---

## Suggestions

- Lead the application with Fenergo agentic document pipeline experience — direct parallel to EIS's insurance document automation use cases
- Address Java gap proactively: emphasise architectural reasoning transferability and rapid framework ramp-up
- Ask in interview: salary band for Ireland-based remote; level of Java used day-to-day vs. Python for AI layer; autonomy level of the agent platform team
- Clarify if the MCP work is greenfield or retrofitting onto existing Java platform stack

---

## Interview Tracker

| Stage | Date | Notes |
|---|---|---|
| Rejected | 2026-05-29 | Java/Spring gap |
| Applied | | |
| Recruiter screen | | |
| Technical interview | | |
| Final round | | |
| Offer / Outcome | | |