AI Development & Engineering — Toby Alexander Rogers

Approach & Core Responsibilities

Using Claude Code as the primary AI coding environment for end-to-end workflow design — from generating and refactoring code to running tests and raising merge requests via standardised slash-command workflows. Atlassian and Glean MCP servers provide live context from Atlassian Jira, Confluence, SharePoint, Teams, and GitLab directly inside each coding session.

Selected Projects

1. Automating BI & Enterprise Solutions workflows

Claude Code · Microsoft Power BI · Power Automate

Objective: double team capacity on BI and platform tasks by offloading repetitive work to an AI coding assistant.

Used Claude Code to build and update Microsoft Power BI dataflows, Power Automate jobs, and related scripts — new reports stood up with significantly less manual effort.
Automated Atlassian Jira/JSM configuration and ticket hygiene — field changes, scheme updates — keeping Atlassian aligned with product and reporting needs without backlog accumulation.

Impact

Delivered ~2× capacity increase for Enterprise Solutions on BI and Atlassian Jira/JSM tasks, unlocking a backlog previously blocked by manual effort constraints.

2. Claude usage analytics

Microsoft Power BI · AI Governance · Claude Code

Objective: give Platform leadership clear visibility into Claude Code adoption across engineering teams.

Co-developed the Dataflow – Platform – Claude Usage Microsoft Power BI dataflow and report, pulling usage data into a standardised model integrated into the Enterprise DataFlows audit suite.
Documented data sources, workspace, and refresh schedule in Confluence to support AI governance reporting.

Impact

Leadership can monitor Claude usage patterns and licence consumption, tied directly to AI enablement initiatives and training programmes.

3. AI-assisted Atlassian workflows via Claude Code + MCP

Claude Code · Atlassian MCP · Atlassian Jira · Confluence

Objective: use Claude Code as an agent over Atlassian to keep code, tickets, and documentation in sync automatically.

Configured Atlassian MCP servers to enable Claude to read/update Atlassian Jira tickets — comments, status transitions, MR links — and edit Confluence pages directly from the CLI.
Applied to Enterprise Solutions work: JSM tickets, Atlassian Jira config changes, and documentation updates all tied into a single agentic workflow.

Impact

Atlassian Jira, Confluence, and code stay in sync automatically — reducing manual bookkeeping and improving auditability of all platform changes.

4. atlassian-mcp — public Python MCP server

Python · FastMCP · Open Source · GitHub

Objective: open-source a production-grade Python MCP server demonstrating the agentic Atlassian pattern delivered at InvestCloud.

Designed and built a read-only Python MCP server exposing curated JQL primitives for Atlassian Jira access via Claude — backend service architecture with FastMCP + httpx + python-dotenv.
Documented architecture, README, demo screenshot, and explicit "not in scope" framing — public artefact at github.com/Toby-Rogers/atlassian-mcp demonstrating senior engineering practice.

atlassian-mcp demo: Claude querying Atlassian Jira via MCP

# Example: a curated JQL primitive exposed as an MCP tool via FastMCP
from fastmcp import FastMCP
import httpx
import os

mcp = FastMCP("atlassian-mcp")

@mcp.tool()
def list_open_tickets_by_component(component: str, limit: int = 10) -> dict:
    """List open Atlassian Jira issues for a given component via JQL."""
    jql = f'component = "{component}" AND statusCategory != Done'
    response = httpx.get(
        f"{os.environ['JIRA_URL']}/rest/api/3/search",
        params={"jql": jql, "maxResults": limit},
        auth=(os.environ['JIRA_USER'], os.environ['JIRA_TOKEN']),
    )
    return response.json()

Impact

Publicly verifiable demonstration of the same agentic Atlassian pattern delivered in production — backend services, AI integration patterns, and product scope discipline as a portfolio artefact.

5. Claude Code Community of Practice & standardisation

AI Enablement · Platform · CoP

Participated in the Platform Claude Code Kickoff and ongoing practices sessions, sharing real Enterprise Solutions scenarios — Microsoft Power BI automation, JSM workflows, Atlassian admin tasks.
Contributions ensure the shared claude-setup framework and skills library reflect actual cross-team production use cases, not just engineering workflows.

Impact

Helps standardise agentic development practices across Platform, making AI-assisted delivery easier to adopt safely and consistently.

6. Glean-driven AI enablement & enterprise context

Glean · Enterprise AI · Cross-system research

Use Glean daily to pull together Atlassian Jira, Confluence, SharePoint, Teams, and Outlook context into a single view when scoping automations or troubleshooting.
BI reports and agentic workflows aligned with the Glean Training & Adoption programme and InvestCloud's broader Enterprise AI strategy.

Impact

AI work is rooted in a secure, permission-aware enterprise context — supporting the wider rollout of Glean and Claude as InvestCloud's primary AI tools.