Chapter 3
PIPER
The smell of frying oil is permanently embedded in my hair.
It doesn’t matter how many times I wash it, or how aggressively I spritz dry shampoo into my roots before class—Dora’s Diner lingers.
Like the ghost of poor life decisions.
Or men who ask for your number while eating a full rack of ribs.
I wipe down table six and mentally re-compile my to-do list for this evening.
Finish debugging the profile matching function, send off the beta version of OptiMatch to Professor Jenkins for feedback, and maybe, just maybe, find five minutes to eat something that wasn’t fried in the same grease as the onion rings.
“Order up!” Marco calls from the kitchen window, sliding a plate toward the edge like he’s in a rom-com. It wobbles. I catch it with a reflex born from too many near-death scrambled egg incidents.
“Got it,” I mutter, swinging it onto my arm and heading for table ten.
An hour later, my shift is over.
I grab my bag and swing by the counter to clock out, peeling off my apron. That’s when I see it—on the floor by table five.
A single green leaf. I roll my eyes.
It must’ve fallen from that guy's plant. Curled and slightly bruised on the edge, like it wasn’t ready to fall off but did anyway.
I crouch, pick it up, and roll the stem between my fingers.
He didn’t leave the plant behind. Of course, he didn’t. He’d carried it out like it was his child.
Still, something about the leaf makes me pause. Makes me think of that boy with the sad eyes and the too-bright grin, sitting alone with a potted friend like it was totally normal.
I smile, just a little.
Then I catch myself.
Nope.
Nope, nope, nope.
Smiling over diner boys is how people end up crying over voicemails and rewatching text threads like forensic evidence.
I don’t do that anymore.
I don’t trust that part of myself.
I trust logic. And metrics. And the elegant, brutal neutrality of code. Specifically, the code that I am designing.
I drop the leaf in the bin, straighten my glasses, and walk out of Dora’s like I’ve never once thought about a boy with a houseplant as a friend.
An hour later, I'm in our apartment bathroom, about to finally wash the diner smell out of my hair when Riya blocks the doorway.
“Absolutely not.”
“I need to shower—”
“You've been avoiding me all day.” She's in full interrogation mode, arms crossed. “Six texts, Piper. Six. You always respond by the third one to tell me to stop being annoying. Something's wrong.”
“Nothing's wrong.”
“You're a terrible liar.” She follows me into my bedroom. “Is this about Miles? Did he text you? I swear to God—”
“It's not about Miles.” I collapse on my bed, still in my diner uniform. “It's worse.”
Her face immediately shifts to concern. “What happened?”
“I'm failing a class.”
“Which—wait, what? You don't fail classes. You get A's while sleep-deprived and subsisting entirely on coffee.”
“Creative Writing.” The words taste like shame. “Thirty-two percent.”
Riya sits beside me, processing this. “The storytelling one? The one you said would be easy?”
“Turns out writing about fictional people with fictional feelings requires understanding actual feelings.” I stare at the ceiling. “Professor Long assigned me a mandatory tutor.”
“Oh, Pipes—”
“And before you say it doesn't matter, it does.” I sit up, pulling my knees to my chest. “Jenkins has a GPA requirement for the research lab. 3.5 minimum. No exceptions.”
Riya's eyes widen. “Shit. The AI lab?”
“The AI lab that's supposed to be my entire senior year. The one with the supercomputer access and the PhD pipeline and the published paper opportunity.” My voice cracks. “The one where my dating app could become actual peer-reviewed research instead of just some bitter girl's side project.”
“You're not bitter—”
“Two hundred people applied, Ry. He picked twelve. I'm the only junior who got in.” I press my palms against my eyes. “Kim Dunn cried when she got rejected, and she has a perfect GPA.”
“Because you're brilliant. Your algorithm work is PhD level—Jenkins said so himself.”
“Yeah, well, apparently I can quantify human emotion but can't write a simple story about it.” I laugh bitterly. “Jenkins wants 'well-rounded scholars.' Technical brilliance isn't enough anymore. You have to score well in a wide range of classes.”
Riya's quiet for a moment. “What about the tutor?”
“Some senior who's supposedly good at 'bridging technical and creative thinking.'” I make air quotes. “I meet them Thursday.”
“So you'll go, you'll learn from them, and you'll pass.”
“It's humiliating. I solve problems that make graduate students cry, but I need help with basic narrative?”
“You need help with one thing. One.” Riya's voice goes firm. “And if some tutor is what it takes to keep Jenkins' position, then you're going to smile and let them explain three-act structures or whatever.”
“You think it’ll be that simple?”
“Of course. My best friend got into the most competitive undergrad research position at this university.” She bumps my shoulder.
She stands. “She can do anything. So, you're going to shower, we're going to dinner with Declan, and Thursday you're going to meet this tutor and be so charming they'll have you passing in no time.”
“I don't do charming.”
“Then do your version. Awkwardly intense. Whatever.” She heads for the door. “But Piper? Don't let pride cost you your future. Jenkins' lab is everything you've worked for.”
After she leaves, I sit with that truth. Three years of perfect grades in every technical class.
Three years of building toward Jenkins' lab, toward proving I belong in AI research despite what Miles made me believe about my emotional limitations.
All of it hanging on learning how to tell a stupid story.
Forty-five minutes later, I’m at Buddha Bowl with Riya and Declan, stabbing at my pad Thai like it personally offended me. The restaurant hums with student chatter, but I barely hear it over the algorithm spinning in my head.
“You’re doing the thing,” Riya observes, dipping a spring roll. “The scary focused face.”
“I’m eating.”
“You’re plotting,” Declan corrects. He’s visiting from Boston again—MIT’s spring break started early. With his wire-rimmed glasses and band t-shirt, he looks like every CS major’s final evolution. “She gets the same look when she’s coding.”
“Speaking of.” Riya steals a piece of my tofu. “How’s Match-y coming?”
“ClearMatch,” I correct automatically. “And it’s... coming.”
Declan leans forward, genuinely interested. He’s good like that—actually listens when people talk about their work. “Remind me how it works? You’re claiming this algorithm can find perfect romantic matches?”
“Not claiming. Proving.” I set down my chopsticks, warming to the topic. “Humans are statistically terrible at choosing partners. You know what percentage of marriages end in divorce?”
“I dunno, twenty-something percent?”
“Forty-one to fifty, depending on demographics. And that’s just the ones that actually end.
Doesn’t count the miserable ones that limp along.
” I pull out my phone, show him the stats I keep bookmarked.
“We let emotions override logic. Get distracted by chemistry, ignore fundamental incompatibilities.”
“Romance is dead,” Riya says dryly.
“Romance is inefficient,” I counter. “We see someone attractive, our brains flood with dopamine and norepinephrine, and suddenly we’re ignoring red flags because they smell nice or laugh at our jokes.”
Declan grins. “You’re really selling the human experience here.”
“I’m being realistic. Computers don’t have hormones. They can analyze compatibility metrics without bias—shared values, communication styles, life goals, behavioral patterns.” I tick them off on my fingers. “Remove emotion from the equation and you get better outcomes.”
“So you’re saying feelings are the problem?”
“Feelings are like…noise in the data. They distract us, stop us from seeing the full picture.”
Riya and Declan exchange a look—one of those couple telepathy moments that I’ve never experienced.
“Okay,” Declan says slowly. “So run us through the app. Hypothetically.”
I light up. “Users input baseline data—age, location, education. Then deeper metrics—career goals, lifestyle preferences, conflict resolution styles. The algorithm weighs compatible traits against complementary ones—”
“Wait.” He holds up a hand. “Test it on us.”
“What?”
“Me and Riya. According to your algorithm, are we a good match?”
Riya snorts. “This should be good.”
I don’t reveal that I have actually used them previously as a case study for my own testing. I had to guess a couple of answers but I think I got pretty close.
“Alright.” I hand Declan my phone. “Input your answers first, then Riya would do the same. This is still the BETA version so ignore the ugly UI. We’ll do the quick version.
I’m also creating an in-depth version which takes much longer to fill out but this will give you a vague idea of what the app thinks. ”
After a couple minutes of them selecting answers, the compatibility is revealed.
“Only 87?” Riya looks mock-offended.
“You’re both introverted but express it differently—she needs solo recharge time, you need parallel quiet activities.
Same values around work-life balance. Similar humor processing—you both appreciate dry wit over slapstick.
” I mentally run through my matrices. “Career trajectories align—tech fields, urban preferences. You argue productively when you disagree. I can confirm that is true.”
“When have you seen us argue?” Declan asks.
“Last visit. The Marvel movie ranking debate. You both stated positions, provided evidence, then agreed to disagree without it affecting dinner. Healthy conflict resolution.”
They’re both staring at me now.
“You’re crazy, you know that?” Declan jokes.
“Yup, anyway, the 13% gap?” I continue. “Riya’s more impulsive with money, you’re a saver. She runs hot, you run cold—literally, your thermostat preferences are incompatible. And your families have different communication styles which could cause friction long-term.”
“Jesus,” Declan says. “You really have been watching us.”
“It’s just data.” I shrug, but something uncomfortable squirms in my chest. “The algorithm would’ve matched you two with 78% confidence on first meeting. Above threshold, which basically means congrats…you guys are a good match...”
“What about that guy you used to hang around with all the time?” Declan asks carefully. “Miles? How would he score?”
“Dec!” Riya chastises him. She knows how upset I’ve been.
My chopsticks pause halfway to my mouth. “I... haven’t run those numbers.”
“Bullshit.”
“Dec! Seriously!” Riya squeals.
“What? Come on. It's a normal question.”
My hand tightens around my chopsticks. I set them down carefully, wiping my suddenly damp palms on my jeans.
“You're so dense sometimes.” She shakes her head.
He’s right. I’ve run them a hundred times, tweaking variables, trying to understand why my brain insisted on wanting someone the data said was wrong.
“42%,” I admit quietly. “We scored 42%. Below threshold. The app knows we would’ve never worked.”
“Because you’re too awesome for him,” Riya says. I roll my eyes.
“I know.” I stop her before she can continue more embarrassing gushing. “Anyway, emotions made me stupid. The algorithm would’ve saved me years of pathetic pining.”
“Or,” Declan says gently, “emotions told you something the data couldn’t measure.”
“Like what?”
“Like maybe you needed that time to figure out what you actually wanted. Maybe Miles was practice.”
“A very long, sad, painful practice.”
“Most important lessons are.”
I push rice around my plate. “The algorithm would've prevented all of it—the heartache, my grades tanking, the humiliation. One compatibility score and I'd have known to walk away instead of wasting years.”
“Would it, though?” Riya challenges. “Or would you have ignored it anyway? You can’t debug feelings, Pipes.”
“Watch me.”
They exchange another look. This is why I need the algorithm—to save people from their own stupid hearts.
“Hey,” Declan says, clearly trying to lighten the mood. “Who would it match me with if I wasn’t taken?”
“Someone who doesn’t leave wet towels on the bed,” Riya mutters.
“That’s the 13%,” I say, and despite everything, we all laugh.
But later, walking home alone while they hold hands ahead of me, I run my own numbers again. 0% chance of meeting someone organically. 0% chance of trusting my judgment after Miles. 100% chance I need this algorithm to work.
Because if I can’t trust my emotions, at least, I can trust the code.
At least, the code makes sense.
Even when nothing else does.