Hi, I'mdigimuza
I help companies
I take money to help companies build and learn AI — hands-on engineering, team training, and strategy that ships. I also use it to run my own day, so I know what holds up past the demo.
What I do, in plain language
Three ways I engage with companies. Pick the one that fits, or combine them.
Build
Hands-on AI engineering
I come in, ship something real, and leave your team with code they can maintain. Prototypes, production features, agent systems — the works.
- ·LLM features & agents
- ·RAG, evals, observability
- ·Prototype → production
Teach
Team workshops & mentoring
Half-day to multi-week programs for engineering teams who need to actually use AI day-to-day, not just talk about it.
- ·Hands-on workshops
- ·1:1 mentoring
- ·Internal AI guidelines
Advise
Strategy & architecture
When you need a second brain before committing. Honest reads on what's worth building, what's a toy, and where to spend the next dollar.
- ·AI readiness audits
- ·Architecture reviews
- ·Roadmap & build-vs-buy
Engagements usually start at 2 weeks and scale up. Remote-friendly, async-first, English + language.
What I work with
The boring part of any AI project is making it survive contact with reality. Here's the toolkit I bring.
Models
- Claude (Anthropic)default for reasoning work
- OpenAI / o-serieswhen the task fits
- Open weightsLlama, Qwen, Mistral on your infra
- Local LLMsOllama for private workloads
Frameworks & patterns
- LangGraphgraph-based agent flows
- Vercel AI SDKshipping in TypeScript
- MCPModel Context Protocol integrations
- Evals + guardrailsthe part demos skip
Languages I ship in
- TypeScriptweb, agents, infra glue
- PythonML, data, glue scripts
- Next.jsUI & full-stack
- SQLyes, still important
Where it lives
- Your cloudAWS, GCP, Azure — your choice
- Vercelfor things with humans
- Fly.io / Railwaylong-running jobs
- On-premwhen the data can't leave
Vendor-agnostic. If your team already uses something else, I'll work inside that stack — not pull you into mine.