Turn complex problems into products that work.
I'm Martín Hounie, a software engineer and Product & AI Engineering Consultant. From Uruguay, for teams anywhere in the world. I help teams understand what to build, use AI where it creates real leverage, and turn ideas into working products and systems.
You probably don’t need "more technology". You need the right problem, solved.
"We know AI matters, but we don’t know where to start."
"Everyone on the engineering team uses AI differently."
"Our coding agents work, but they’re inconsistent."
"Our information is spread across too many tools."
"We have an idea but don’t know what to build first."
"Our current product is getting hard to change."
"We need senior technical judgment, not a full-time hire."
"We need to build something."
Technology should amplify judgment, not replace it.
Software engineering is changing. Increasingly, value comes not just from writing code, but from understanding the problem, defining the right product, creating useful context, connecting systems, choosing tools, directing agents and maintaining quality. That is the layer where I keep focusing more of my work.
How I can help
Starting with hourly consulting is the easiest way to work together — not the ceiling of the value.
See all services →Product & Technical Consulting
Discovery, architecture, feasibility, roadmap, build vs buy and second opinions for ambiguous decisions.
AI Opportunity & Workflow Audit
Where AI and automation create real value — no random experimentation. Opportunity map + 30/60/90-day roadmap.
AI Workflow & Team Enablement
A coherent, practical way of working with AI for your actual team. Not generic training — workflow design.
Agent-assisted Engineering Workflow Sprint
Reliable agent-assisted workflows: context, instructions, skills, tools, review gates and feedback loops. Agentic and harness-engineering practices.
AI Workflows & Automation
Prototypes and systems combining LLMs, APIs, company data and human approval. Human in the loop.
Product Engineering / MVP
Build and improve digital products. Frontend, React, TypeScript and maintainable product UX.
Where I apply AI
Applied AI for real teams and real products. Practical, reviewed and human-controlled — no hype.
How I think about AI →AI for Engineering
Moving teams from casual AI usage to intentional AI-assisted engineering: coding agents, repo instructions, context engineering, review loops and automation. Controlled and reviewed execution — not autonomous development.
Agentic Engineering & Harness Design
A model alone is not enough. Useful systems need context + instructions + tools + data + permissions + feedback + evaluation + human oversight. I help design those systems.
AI for Knowledge & Operations
Knowledge lives scattered across documents, email, PM tools, code and APIs. I design workflows where AI finds, summarizes, compares and prepares actions — human in the loop.
AI-native Product Building
Using AI as leverage to build real products: from problem and discovery to system design, build, review and ship.
Selected work
4+ years with US clients —healthcare, security and more— inside engineering teams. Confidentiality keeps them unnamed.
He consistently delivered clean, well-structured code and brought a thoughtful approach to solving problems. Beyond his technical skills, he stood out for his great attitude and the positive energy he brought to the team — personable, collaborative and easy to work with.
Experience
View full CV →Web product engineering with US teams and enterprise environments. Cross-functional work on large, evolving products; engineering judgment and delivery.
Digital product, reservations, payments and operations for a real trailer-rental business.
Web and mobile development with a US team. React and React Native for digital products.
Road-safety startup. Paused after receiving backing from the Uruguayan innovation ecosystem (ANDE, Incubaelectro, Antel, LATU, MIEM, INITIUM, UM, UNASEV).
The human part matters too.
I’m excited about AI, agents and all the leverage we can build around them. But products are still born from people who understand problems, discuss ideas, make decisions and trust each other.
I especially enjoy that part of the work: listening, explaining, untangling complex problems and helping a team move forward with clarity.
I want the outcome to be good. And I also want getting there to be a good experience.
I’m interested in the intersection of technology, product, business and people.
I enjoy understanding messy problems and turning them into simpler systems. I hold a Computer Engineering degree with a diploma in Business Management from Universidad de Montevideo, and I started my career very close to implementation.
Over time I got more and more interested in everything around the code: understanding which problem is worth solving, defining product, communicating decisions, connecting people and systems, and now designing better ways to work with AI and agents. I also genuinely enjoy working with people: explaining ideas, listening to different perspectives and creating an environment where it’s easy to collaborate.
AI Lab
Experiments in agents, automation and AI-native product engineering.
Agent-ready Repository
Experiment / Coming soonConnected Knowledge Assistant
Experiment / Coming soonAI Product Discovery Workflow
Experiment / Coming soonMulti-agent Product Build
Experiment / Coming soonHow I think
Solve the problem, not the ticket.
A perfectly implemented feature is useless if it solves the wrong problem.
Understand before automating.
Automation magnifies both good and bad processes.
AI should create leverage.
Using AI is not the objective. A better outcome is.
Humans own judgment.
Agents can execute. People remain accountable.
Context is infrastructure.
A good AI system depends on what it knows, what it can access and how clearly it understands its environment.
Ship and learn.
Reality produces better information than speculation.
Ways to work together
Company engagements are scoped independently and can be executed through Mahomarino Technologies SAS.
Indicative prices. Taxes not included (VAT or the applicable tax by country).
Tell me the problem.
No pitch deck required. Tell me the problem and we’ll figure out whether I can help.
Book a 30-minute intro call →A conversation, not a sales pitch. Share the context and we’ll see if it makes sense.
Full contact page →Thanks — I’ll be in touch soon.
Got your message. I’ll reply within 48 hours.