I Ran a Manager-Worker Agent Team on Real Work. A Single Session Beats It.
I ran a manager-worker agent team on 67 real tasks. A single session would have done the work better. The failure modes are structural.
BLOCKHEAD CONSULTING
Your team uses AI tools. I'll show them how to ship with them.
I'm a senior systems engineer who builds and ships production code with AI tools daily.
10+
Years Systems Engineering
Go + Python
Core Languages
Ships AI
Practitioner First
Where I Add Immediate Value
Your developers use AI tools but results are inconsistent
→
Structured AI workflows with persistent context and documented conventions
→
Every developer on your team produces consistent, high-quality output with AI assistance
You're spending on AI tools but can't measure ROI
→
Instrumented development workflows that capture real output (lines shipped, features delivered, time-to-merge) alongside AI tool spend
→
Concrete metrics for leadership: what was built, how fast, at what cost
AI agents lose context between sessions, causing rework
→
Persistent agent workspaces with CLAUDE.md conventions, structured planning per session, and context handoff between agents
→
Fewer tokens burned on re-explanation, less rework, faster iteration cycles
Public Artifacts
Public work that demonstrates the practice. Click through to verify directly.
Services
AI/LLM Consulting
- AI Workflow Architecture for Engineering Teams
- Persistent Context for Long-Running Agents
- Instrumented Development Pipelines
- Senior Engineer Coaching for Agentic Workflows
- Embedded Engineering with Your Team
Crypto Infrastructure
- Decentralized Social Media Platforms
- Cross-Chain Systems
- Custodial Wallet Systems
- Chain Listeners/Executors
- NFT/Token Platform Development
- Performance Optimizations
- Chain Migrations
Lance Rogers
Backend engineer with 10+ years shipping production systems across fintech, security, blockchain, and developer tooling.
Currently building AI execution infrastructure: LLM orchestration, MCP, multi-agent systems, and evaluation tooling.
Known for owning ambiguous systems end-to-end, from architecture through operations.
Technical Expertise
Core Languages
Go • Python • Solidity
Blockchain
Ethereum • Polygon • DeFi • Smart Contracts
AI Engineering
OpenAI • Claude • Agent Frameworks • RAG
Core Languages
9+ years developing production systems in Go and Python, with specialized expertise in Solidity for blockchain applications.
Experience
- Go: 7+ years building microservices, APIs, and blockchain infrastructure
- Python: 9+ years in fintech, data pipelines, and automation systems
- Solidity: 4+ years developing and deploying smart contracts
Technologies & Frameworks
- Go: Gin, Gorilla Mux, gRPC, concurrent programming, memory optimization
- Python: Django, Flask, FastAPI, NumPy, Pandas, async programming
- Solidity: ERC standards, gas optimization, security patterns, upgradeable contracts
Key Achievements
- Authored claude-code-go SDK (first Go wrapper for Anthropic's Claude Code)
- Built Go microservices handling 15M+ transactions/month at Mythical Games
- Deployed audited smart contracts for gaming and marketplace applications
Blockchain
4+ years architecting blockchain systems from smart contracts to full-stack dApps, with deep expertise in multiple networks and protocols.
Experience
- Smart contract development and deployment on multiple networks
- Cross-chain bridge architecture and implementation
- DeFi protocol development and integration
- NFT marketplace and gaming platform development
Technologies & Frameworks
- Networks: Ethereum, Polygon, Hyperledger Besu, Layer 2 solutions
- Tools: Hardhat, Truffle, Web3.js, ethers.js, OpenZeppelin
- Protocols: ERC-20/721/1155, DeFi primitives, cross-chain bridges
Key Achievements
- Discovered critical Polygon Edge bug before mainnet launch at Mythical Games
- Built NFT marketplace infrastructure handling millions in transactions
- Architected P2P precious metals exchange with secure escrow (ShinySwap)
- Led blockchain meetup teaching 8000+ community members
AI Engineering
Specialized in production AI systems, from LLM orchestration to multi-agent frameworks, with focus on practical business applications.
Experience
- LLM integration and optimization for enterprise applications
- Multi-agent system design and orchestration
- RAG (Retrieval-Augmented Generation) implementation
- AI workflow automation and productivity tools
Technologies & Frameworks
- APIs: OpenAI, Anthropic Claude, Ollama for local deployment
- Frameworks: Custom agent orchestration, MCP (Model Context Protocol)
- Tools: Vector databases, embedding systems, prompt engineering
Key Achievements
- Created Guild Agent Framework for multi-agent orchestration
- Authored first Go SDK for Anthropic's Claude Code CLI
- Built YouTube summarizer with local LLM processing
- Consulting clients on AI workflow integration and productivity gains
Latest Technical Insights
Your Agent's Memory Should Be Files in Git, Not Rows in a Vendor's Database
Agent memory should be files in git, next to the work they describe, not rows in a vendor database.
Quality Gates for Agent Work: Decisions, Not Formalities
A gate an agent waves itself through is not a gate. Quality gates for agent work have to be decisions against artifacts.
Ready to fix what's slowing you down?
Most scaling problems have a 2-week solution and a 6-month solution. Let's find yours.
No sales calls. Just an engineer who gets it.