It is well known that cells can respond to diffusible molecules but it is less well known that they are also able to respond to patterned surface topographies. If we are able to understand how cells respond to these patterns, we can rationally design the surface of medical implants for optimized functionality. To unravel the secret Braille language of cells, we have designed and engineered the TopoChip platform, a library of surface topographies reproduced onto polymeric surfaces. Using high-content imaging, we are able to analyze the response of cells to thousands of different surface patterns simultaneously. Thus, we have found surfaces which induce expression of the early osteoblast marker alkaline phosphatase (ALP) in mesenchymal stromal cells to levels similar to that induced by classical osteoblast inducers such as dexamethasone. In addition to ALP staining, we have also stained the actin cytoskeleton and the nucleus, and using CellProfiler software, we have extracted nearly 300 morphological features for each cell on the TopoChip. Using machine learning algorithms, we are now able to predict ALP expression based on cell morphology alone, and further experiments are in progress to investigate a possible correlation between the actin cytoskeletal organization and ALP expression. In conclusion, using a high-throughput screening approach, we can now start to unravel the secret language of surface topographies and apply the hit surfaces to improve medical implants.