Rewards Hub 
She started coding on Friday at 9 PM with a cup of coffee and a theory about range-bound markets.
By Sunday afternoon, she had a functioning grid trading bot running on Phemex's API, and she hadn't left her apartment once.
This is how she spends most weekends.
The Market Doesn't Care About Your Feelings
Her friends trade on "vibes." They talk about being "bullish" or "bearish" like these are real analytical frameworks. They check charts obsessively, convinced they can feel where the price is going next.
She doesn't feel anything. Markets are systems. Systems follow patterns. Patterns can be coded.
She's been trading since 2021, but she stopped manually trading in 2022. Not because she was bad at it — she was fine — but because she realized emotional discipline is hard. Code doesn't have emotions.
Why fight your psychology when you can just remove it from the equation?
May 2021: The Bot That Broke Everything
Back when she thought she was smarter than everyone else.
May 19, 2021. Bitcoin dropped from $43K to $30K in about four hours. She was at work, in a meeting about some database migration, when her phone started going off. Error notifications. One after another.
She excused herself, went to the bathroom, pulled up her terminal.
Her momentum bot was liquidating itself in real-time.
The logic was sound. Buy breakouts, sell breakdowns, trailing stops. Worked great for two months. Up 40%. She'd shown it to her coworker like "see, this is how you actually trade."
But when volatility spiked and price started whipsawing, the bot kept buying breakouts that immediately reversed. Bought at $38K, stopped at $36K. Bought at $39K, stopped at $37K. Over and over. Seven times in an hour.
By the time she killed it manually, down 35%.
She sat in her car after work just staring at the dashboard. The bot didn't malfunction. It did exactly what she coded it to do. The code was just wrong for that situation.
Pulled up Twitter. Everyone's bots were dying. "My algo just got rekt." "Turns out my strategy only works in bull markets lol." At least she wasn't alone in being an idiot.
Except three platforms had worse problems. Order execution delays. API timeouts. One liquidated people at prices that didn't match anywhere else because their system couldn't handle the load.
Phemex executed everything. Her stops triggered when they should. The losses were her fault, not the platform breaking.
Small consolation when you just lost 35% because your clever code turned out to be not that clever.
2022: Watching the Smart Money Fail
May 2022. Luna imploded.
She watched it happen in real-time on Twitter. An algorithmic stablecoin designed by people with PhDs. Game theory, arbitrage mechanisms, death spirals prevented by math.
Except the math was wrong. Or the assumptions were wrong. Or both. $40 billion gone in 48 hours because the algorithm accelerated the collapse instead of preventing it.
Her friend — also an engineer, also thought he was smart — lost $80K in UST.
"The mechanism design made sense. How did it not work?"
Because you can't code around human panic. Because edge cases you didn't think of will kill you.
She was building trading systems while watching other people's trading systems implode. Celsius couldn't handle withdrawals. Three Arrows Capital was just overleveraged gambling. BlockFi, Voyager, all the "algorithmic" platforms — just badly managed risk.
Then FTX. November. The exchange run by "quant traders" who supposedly understood risk management better than everyone else. Their algorithm for managing customer funds was just fraud with extra steps.
She started adding more circuit breakers to her bots. More "if weird shit happens, stop everything" logic. Made less money. Survived.
The Weekend Project
Bitcoin stuck between $98K and $103K for two weeks. Perfect for a grid bot.
The idea is simple: buy orders below current price, sell orders above. Price bounces around, you capture the spread.
Simple idea, annoying execution. She spent Friday night writing the placement logic, realized her rebalancing rules were garbage, rewrote them. Spent an hour debugging why her websocket kept dying before realizing she forgot to send a heartbeat.
It's always something stupid.
Around 2 AM she ordered pad thai and kept coding.
Saturday morning, paper trading mode. First bug: orders placing outside her range. Fixed. Second bug: position sizing wrong. Fixed. Third bug: fat-fingered a variable name, took 45 minutes to find because of course it did.
Found eleven bugs total. Ran clean for two hours in paper mode. Good enough.
Switched to live. Immediately crashed. Forgot to handle the exchange's minimum order size.
Fixed. Restarted. Watched it for an hour. Everything executed clean.
Closed her laptop, went for a walk. If it breaks, it breaks.
APIs That Actually Work
She'd tried building bots on other platforms. Always a disaster.
Rate limits that triggered randomly. REST endpoints timing out during volatility — exactly when you need them. Websocket feeds just deciding to stop sending data. Good luck figuring out why from their documentation.
And getting accurate margin data programmatically? Half these platforms don't even expose it properly. You're supposed to just trust their liquidation engine.
She'd lost count of how many times a bot failed because the exchange's API was garbage, not because her code was bad.
Phemex's API was just... functional. Docs matched the endpoints. Rate limits made sense. Error messages actually told you what went wrong instead of "bad request."
And the Unified Margin system meant she didn't have to manually move collateral around. Her whole account backed every position. For grid trading, that meant she could run 18 levels instead of 8 with the same capital.
She set up 18 grid levels between $98,400 and $102,600. Each level: 0.03 BTC. Stop below $96K. Close everything above $105K.
Saturday afternoon, after fixing three typos and one race condition that actually mattered, the bot went live.
Watched it for an hour. Everything worked.
Then she stopped watching because staring at logs doesn't make code better.
Sunday Morning
Woke up, grabbed her phone.
Fourteen trades overnight. Eight buys on dips, six sells on bounces. Net P&L: +$410.
Not life-changing. Just the system working while she was asleep.
She didn't wake up at 3 AM to manually trade. She didn't miss anything because she was making breakfast. The bot just ran.
By Sunday evening, 34 trades total. +$920. Not some moonshot, just consistent execution.
She checked the logs twice looking for issues. Found none. Everything clean.
Code working as intended is better than the money.
Sunday Night Doubt
Late Sunday, scrolling Twitter. Someone posted a 40x return on some memecoin. Comments full of "just bought more" and rocket emojis.
Her bots made $920 this weekend. This random person made $120K clicking buy.
Happens every cycle. Manual traders with no system, no risk management, no code — just vibes and luck — making 100x while she grinds consistent returns with sophisticated infrastructure.
What's the fucking point of all this if some idiot with no code makes 100x?
Her ex used to say shit like this. "You spend your whole weekend coding to make $900? Just buy Bitcoin."
Yeah. Or buy Bitcoin at the top and lose 60%. Or buy some shitcoin that dies. Or panic sell at the bottom because humans are terrible at this.
The systems don't make you smarter. They just remove the part where your emotions destroy you.
Still. Watching someone make six figures on a memecoin while she's debugging websocket connections at 2 AM... makes you wonder if you're doing it wrong.
Three Years In
Three years building these systems. One lesson: strategy is easy, execution is everything.
Doesn't matter how good your logic is if the exchange crashes during volatility. Your arbitrage bot is useless if API rate limits kick in when spreads widen. Grid strategies fail if you can't get accurate margin data.
She runs six different bots on Phemex now. Grid strategies, DCA scripts, funding rate stuff. Not all winners every week. But they execute reliably because the infrastructure works.
Phemex's API uptime is basically perfect. Orders go through. Data feeds don't die. Margin calculations are correct. Two years running these bots, zero failures caused by the API itself.
After watching Luna's algorithm fail, after seeing FTX's "risk management" exposed as fraud, after her own bots breaking on shitty infrastructure — clever code means nothing if the foundation is broken.
Or maybe: none of this shit matters if the exchange goes down.
Still Building
Day job: software engineer at a fintech company. Nights and weekends: building trading bots because apparently coding all day isn't enough.
Her portfolio isn't huge compared to friends who hit big on memecoins. But it's consistent. They have crazy wins and brutal losses. Her account just grinds up. Some weeks green, some red. The bots keep running.
People ask for trading advice sometimes. She tells them: "Don't try to predict the market. Design a system that can survive it."
Most people don't want to hear that. They want hot tips, not Python tutorials.
Fine. Less competition.
The Satisfaction
There's something good about waking up and seeing your code executed perfectly overnight. Not exciting. More like... it worked how it was supposed to.
Logic was sound. Code was clean. Infrastructure held.
Her grid bot is still running. Bitcoin's still chopping between $98K and $103K. As long as it stays in range, the bot keeps farming. Breaks out? Positions close automatically, waits for the next setup.
She doesn't need to watch it.
She's already working on the next thing. Some liquidity gap stuff around funding rate resets. Early backtests look okay. Probably running by next weekend.
Assuming she doesn't spend four hours debugging some dumb typo again.
Which she probably will.
To uphold our uncompromising first principle — the absolute protection of our users' privacy and asset security — the characters and narratives in our 'Legends' series are artistic creations, forged from the real actions and behaviors of millions of users within our community. They are the embodiment of the Phemex community's spirit. Behind every legend, we see the imprint of you.
The greatest systems aren't just built on weekends; they are proven in the arena.
This anniversary, we're rewarding intelligence.
Phemex is celebrating 6 years of reliability with a $6,000,000 Trading Competition—a laboratory where code meets capital and logic is tested against the chaos of the market. It's not about lucky trades; it's about proving your system has a quantifiable edge.
Deploy your strategy.
