Practice for quant and AI engineering interviews, round by round.
Pick your role
Each role is its own set of rounds, and your practice follows it.
Quant
A quote board reading bid 13, ask 15, with the spread marked between them, and two playing cards face down in front of it.
A backtest on graph paper: the line climbs in sample, then sags out of sample, in orange, under a reading glass.
An order book drawn as queues: asks above, bids below, each price level a line of orders in arrival order. An incoming buy in orange takes the first ask in line.
AI engineering
A retrieved page with one passage marked in orange, a chat answer that cites it, and an eval set on a clipboard with ticks and one cross.
A training chart: training loss keeps falling, validation loss falls then turns up. An orange dot marks its lowest point, at epoch 6, and the checkpoint saved there is lifted out of the row below.
One GPU's 80 GB of memory, drawn exploded: the model's weights at the bottom, a slab of KV cache for each request in flight, dashed free space above, and the next request's cache in orange, waiting to drop in.