The AI field is enormous. Your next move can be clear.
See the whole landscape, find the branch that fits you, and turn learning into public proof—from Medellín to global work.
Find a credible first direction
This is a hypothesis, not a permanent identity. Test one primary route and keep one adjacent skill.
The AI capability map
Eight durable branches organize the field. Tools change quickly; these underlying problems persist.
Opportunity explorer
Filter by capability, coding depth, proof speed and market reach. A role title matters less than the evidence you can produce.
What the industry is emphasizing now
The 29 AI Engineer World’s Fair tracks are normalized into six learner decisions. Conference attention is a signal—not a curriculum or vendor ranking.
A low-cost learning system
Do not collect courses. Advance when you can explain the level, build its evidence and identify your remaining gaps.
Project ladder
Each project should leave behind an artifact another person can inspect: code, evaluation, architecture, decisions and measured results.
Medellín launchpad
Use one builder community, one language community, one platform community and one institutional feed. Attendance becomes valuable when it produces relationships and visible work.
Productized offers for US and global buyers
Sell a bounded outcome before selling open-ended hours. Use bilingual context, overlapping time zones and domain knowledge as advantages—not substitutes for evidence.
Production and governance checklist
Reliable AI is not a clever prompt. It is a controlled system with evidence, permissions, traceability, human authority and recovery.
Ecosystem library
Representative tools are grouped by capability, not ranked. Official links open in a new tab. Reverify current products, pricing and availability.