Claude Code Workshop: For Engineering Teams
Your engineering team puts Claude Code into the daily development loop. Spec writing, code review, custom MCP servers, agentic refactoring — end of workshop, a live Claude Code setup in your own repository.
Duration
2 days (12 hours)
Format
Hands-on workshop · live code in your IDE · 6-12 seats
Starting at
$850/seat
Claude Code is Anthropic's terminal-based AI engineering tool: it runs in an agentic loop, thinks in plan mode, and binds to your codebase through MCP. Used right, it multiplies senior engineering output 2-3x; used wrong, it generates poor code and doubles review load. This 2-day workshop sets your team up with senior-level patterns: context engineering with CLAUDE.md, custom MCP server build, slash commands and hooks, subagent-driven code review, production guardrails. Comparable depth to Frontend Masters' Claude Code Deep Dive, but applied live in your real repository.
Tools we'll use
Who this is for
- 01Software engineering teams and tech leads (6-12 seats)
- 02Engineering managers and VPs scaling AI productivity
- 03Teams who tried Claude Code, felt 'this is a toy,' and stopped
- 04Platform teams wanting to spread senior productivity to junior engineers
Modules
Inside the curriculum
01Claude Code mental model
+
CLI, agentic loop, plan mode, permission mode, model selection (Sonnet 4.6 vs Opus 4.7); what it's for, what it's not.
02Context engineering: CLAUDE.md and good specs
+
Telling the agent about your codebase; example-driven guidance; multi-turn thinking; spec → plan → diff discipline.
03Custom MCP server build (live)
+
Build an MCP server for your Postgres, Linear, Jira, or internal services; live deploy and share with the team.
04Skills, slash commands, hooks, subagents
+
Custom slash for repeated work; pre/post hooks as guardrails; subagents for parallel work; code review automation.
05Production guardrails and team integration
+
CI-integrated review; secret scanning; human approval gates; story-point/sprint productivity measurement playbook.
Outcomes
What you'll walk away with
Your engineering team puts Claude Code into the daily development loop. Spec writing, code review, custom MCP servers, agentic refactoring — end of workshop, a live Claude Code setup in your own repository.
- 01Company-specific CLAUDE.md and code-style guide (produced in the workshop)
- 02Working custom MCP server (bound to your real services)
- 03Subagent and hook configuration (.claude/ directory, shareable)
- 04Productivity measurement playbook (per story-point/sprint)
- 0530 days email support + 2 follow-up calls
Rotterdam · IstanbulFOUNDER
Founder & Lead Advisor
Burhan is the founder and lead advisor at Leftflow. He runs every engagement — picks the team, owns the timeline, sits in every working session. No account managers, no handoffs.
Before Leftflow, he built ARspar — an AI visualization platform — to 70+ enterprise clients and €1M+ in revenue before its acquisition by Floorplanner in May 2024. He was named among the Netherlands' Best Under-25 Entrepreneurs in 2022.
- Best Under-25 Entrepreneur NL · 2022
- ARspar acquired by Floorplanner · 2024
- €1M+ revenue generated
- 70+ enterprise clients delivered
Frequently asked
Claude Code Workshop: For Engineering Teams
How does it differ from Cursor, Copilot, Windsurf?
+
They are complementary. Claude Code is agentic — it runs long tasks in the background, plans, and modifies across the codebase; Cursor/Copilot are inline suggestion-based. We show how to combine them for your stack.
Which languages / frameworks?
+
Language-agnostic. We work on your real repository (TypeScript, Python, Go, Java, Ruby — whatever).
Too basic for senior teams?
+
Modules adjust to skill matrix. Senior-heavy teams get MCP development + subagent architecture + production guardrails as the main thrust.
What productivity gain should we expect?
+
Anthropic's own benchmark is 30-50% gain by seniority. Across our clients, by week 6 we observe 25-40% improvement per story-point.
Which program is right?
On a 30-minute diagnostic call we assess your situation and recommend the right program — or tell you honestly if none of them fit.
Book the call→