Twin Cities .NET User Group

From Solo Prompts to Autonomous Teams

What happens when you move past standard AI code completion and deploy an entire fleet of specialized agents working in sync?
We kick off this session with a live agentic development demo: watch four specialized Claude agents collaborate in real time—handling architecture, frontend design, backend logic, and automated testing—to build and launch a full-featured web application from scratch.

Key Session Takeaways

• The Claude Architecture Unpacked: A deep dive into the core primitives powering Claude’s ecosystem—from Custom Skills (tailoring agent behavior and team coding standards) to Plugins & Tools (connecting Claude directly to your APIs, databases, and dev stack).
• Unifying Team Workflows Across Surfaces: Learn how codifying skills across your organization creates a shared AI baseline as we tour Claude’s primary developer interfaces:
• The Desktop App: High-level project planning, architectural design, and visual collaboration.
• The VS Code Extension: In-editor contextual refactoring and real-time inline coding assistance.
• The Terminal (Claude CLI): Where high-volume execution happens—orchestrating autonomous multi-file edits, build loops, and git workflows directly from the command line.
• Reframing “Token-Maxxing” for Leadership: Why raw token consumption alone is a dangerous metric, and how to optimize context window efficiency without inflating cloud bills or getting stuck in prompt-and-pray loops.
• The C-Suite Framework: Measuring Real Engineering ROI: Moving beyond vanity metrics like “lines of code generated” or “active seat count.” We introduce compound metrics—such as Tokens per Story Point and Tokens per Merged PR—giving CTOs, CEOs, and engineering leads an actionable baseline to evaluate true developer velocity, AI efficiency, and code quality.

Who Should Attend
Software engineers, tech leads, CTOs, and engineering managers looking to level up from basic pair-programming tools to autonomous agentic workflows, standardized organization-wide AI integration, and business-focused ROI metrics.

Please attend for a chance to win a license from our partners at JetBrains.
This will be an online-only event, with the link below!

We’ll start at 5:30 PM CT, but feel free to grab a beverage and join a bit earlier!

AI is changing how software is developed, moving beyond code completion and chat-based assistance toward autonomous agents capable of planning, implementing, testing, and validating software.

In this session, we'll explore agentic engineering using Cursor, examining how AI agents can participate in the entire software development lifecycle—not just writing code, but translating business requirements into working, verified software.

Through a live demonstration, we'll walk through a multi-agent development workflow that starts with a meeting transcript and progresses through requirements analysis, implementation planning, human approval, code generation, automated testing, and quality verification.

We'll also explore how Cursor's agents, subagents, skills, and development tools work together, along with the engineering guardrails needed to maintain code quality, architectural consistency, accessibility, and design standards.

Please attend for a chance to win a license from our partners at JetBrains.
This will be an online-only event, with the link below!
We’ll start at 5:30 PM CT, but feel free to grab a beverage and join a bit earlier!

TOKENOMICS: From Sloppy Usage to Sustainable Scale Before the AI honeymoon ends.

> “If the subsidies covering your AI usage were cut tomorrow, could your team keep delivering, or would your workflows collapse under the true cost?”

What we’ll cover

1. The Reckoning: Your AI bill doesn’t need a price hike to get worse.
2. Deterministic Design: Stop renting intelligence for rules you already know.
3. Context Engineering: Your context window is not a landfill for agent retries.
4. Model Routing: Stop paying frontier rates for routine work.
5. Hybrid AI: Stop letting one vendor hold the off switch.
6. The Custom Harness: Build true gating, open source routing, and shims.

The ultimate test
Every optimization faces the same test: the Iron Triangle of Quality, Speed, and Cost. Does it deliver the required quality, at an acceptable speed, for a cost the business can sustain?
What you’ll take away
A practical approach to finding waste, controlling execution, and maintaining options when a provider changes the deal.

Please attend for a chance to win a license from our partners at JetBrains.
This will be an online-only event, with the link below!
We’ll start at 5:30 PM CT, but feel free to grab a beverage and join a bit earlier!