What an AMM is, in 30 seconds.
An automated market maker is a smart contract that holds two tokens in a pool and lets anyone trade between them. Instead of an order book with bids and asks, it uses a simple rule: as one side of the pool gets bought, that side gets more expensive, and the other side gets cheaper. The price comes out of the math, not from a counterparty.
When you provide liquidity, you deposit both tokens into the pool. Traders pay a small fee on every swap. That fee gets shared across all liquidity providers in the pool, in proportion to how much liquidity they contributed.
That's the whole product. The complexity comes from one question: where in the pool does your money actually live.
Old-school AMMs spread your money everywhere.
The first generation of AMMs, like Uniswap v2, took your deposit and spread it across the entire price range. From zero to infinity. Most of that range never gets touched. If SOL is at one hundred and fifty dollars, almost no one is trading it at five dollars or five thousand. So most of your capital sits there, doing nothing, earning no fees.
It was simple, but it was wasteful. You needed enormous pools to keep slippage low.
Concentrated liquidity changes that.
Uniswap v3 and Meteora DLMM introduced the same idea: let LPs pick a price range. Instead of spreading your deposit from zero to infinity, you concentrate it around the current price. If SOL is one hundred and fifty dollars, you might say: I want my liquidity active between one hundred and forty and one hundred and sixty.
Inside that range, your capital does the work of much more. Every dollar you deposit might earn the fees of ten or twenty dollars in an old-school pool. That's the whole pitch.
Meteora DLMM goes one step further by splitting the price range into discrete bins. Each bin is a tiny slice of price. You choose which bins to fill. Below the active bin you're holding the quote token, above it you're holding the base token, and at the active bin you're earning fees on every swap that touches it.
The catch: tight ranges go out of range.
Concentration is a tradeoff, not a free lunch. The tighter your range, the more fees you earn while price sits inside it. The tighter your range, the faster you fall out of it when price moves.
When you're out of range, you stop earning fees entirely. Your position turns into a one-sided bag: all of one token, none of the other. You either wait for price to come back, or you reshape the position to a new range.
Wide ranges feel safer because they stay in range longer. But they earn less per dollar. Tight ranges earn more per dollar, but they need attention. This is the core LP tradeoff and there is no way to engineer around it.
Why this is hard to do by hand.
If you set a range and walk away for a week, on a volatile pair the price will leave your range. You'll come back to a stale position, no fees earned, and a portfolio that drifted. If you set a range and watch it like a hawk, you'll spend every spare minute deciding whether to reshape, paying gas, paying slippage, second-guessing yourself.
Most people who try to LP manually either go too wide and earn very little, or pick a tight range and forget about it and find out a week later that they've been out of range the whole time.
This is why agents matter.
An agent is a small piece of automation that watches the price, watches the volatility, and decides when the cost of reshaping is worth less than the fees the new range will earn. It does this every block, not every weekend. It does it without ego, without hope, without checking the chart for confirmation.
The rest of Academy is about how Lyra agents do that work. Concentrated liquidity is the engine. Agents are the driver. You can't have one without the other and expect a good ride.