Single Hearts

My hands trembled slightly when I hit save on the final commit. It was nearly four in the morning. The only light in the room came from two laptops and a cheap desk lamp whose bulb flickered like it was as exhausted as I was. The air smelled of stale coffee, cold pizza, and the faint mold the landlord had ignored for months.

This was our “office”. A twenty-eight square meter room I had rented in the same building where I lived, one floor down from my actual apartment. Calling it an office made it easier to explain to people.

It had my old university desk, two chairs, a corkboard covered in printed screenshots and handwritten notes, and a copy of a research paper from a lab in Zurich. The rest of the walls were covered in yellow Post-its filled with formulas, ratios, and half-finished sketches. That Zurich paper had started everything. Or rather, it had given a name to something I had already suspected.

Alex sat hunched in my old gaming chair, fingers moving methodically across the keyboard. He was the only one who had never laughed at me. We had met in college. He was the quiet genius in the back row who solved problems before the professor finished explaining them.

After graduation, when I told him I had a crazy idea and almost no money, he just shrugged and said, “Okay. When do we start?”

Our first attempt had been an intelligent maps app. Something that calculated emotional distances between places based on reviews and behavioral data. It failed completely. Nobody wanted to know how depressing their commute was. But in the middle of that failure, staring at heat maps and convergence points at two in the morning, something had occurred to me. Distance was never a straight line. People thought it was. It never was.

That thought took six months of reading to become a theory. The research was there if you knew where to look. Studies on facial symmetry. On subconscious recognition. On how we register similarity in bone structure before we’ve consciously processed anything. The Zurich paper was the clearest of all of them. It argued that we are drawn to faces that match our own not out of vanity but out of something much older. Recognition. The sense of a self that can be understood.

The theory had a name. Facial imprinting. The idea that we carry a template, built early and built deep, of what a face should look like. And that template, more often than not, looks like us. Not the way a mirror does. The way a cousin does. Someone who shares the same underlying geometry without sharing the same face.

Before I had written a single line of code, I had tested the theory on myself. I opened my phone. I found photos of my four ex-boyfriends. I ran rough measurements on each of them. Not the things people usually talk about when they talk about attraction. Not whether someone had full lips or sharp cheekbones. The distances between things. The ratio of upper face to lower face. The space between the base of the nose and the upper lip. The jaw angle. The internal logic of a face.

The results were not subtle.

The one I had dated longest, almost two years, the one I had been most certain about, scored highest. Not because we looked alike. Anyone seeing a photo of us would have said the opposite. But the proportions matched. His orbital angle. The ratio of his nose to his mouth, within three percent of mine. The geometry underneath was the same even when the surface wasn’t.

The one I had connected with least scored the lowest. By a significant margin.

I sat with that for a long time. Then I called Alex.

---

“I’ve finished the new weights,” I told him now, my voice rough from hours of silence. “Added the micro-asymmetry indexing, the nasal-labial ratio adjustment, and the lighting correction for artificial sources. Also the mandible angle variance at different head tilts.”

He nodded without looking up. “Running integrity tests.”

I stood and stretched, my back aching. Our bank account was dangerously low. I could afford Alex’s modest salary, the rent on this room, and the server costs. My parents still helped occasionally. Gas money. Something for the car when it made sounds it shouldn’t. I hated needing it. But without it, none of this would still be running.

The site had been live for eight months. It wasn’t thriving. It hadn’t died either. A red banner had sat on the homepage for two weeks. Major algorithm update coming. Get ready for matches that feel real. Most users thought it was marketing. I knew what it actually was.

A few minutes later he looked up from the screen. “Done. Everything looks good.” He paused. “We should run a real test. Not the validation set. An actual face.”

“Sure,” I said. “We can pull someone from the existing profiles.”

“No.” He paused. “You. Real conditions, bad lighting, no filter. Like any other user would.” He was already picking up his phone. “It’s the only way to know if it holds.”

I looked at him for a moment. It was a reasonable argument. Methodologically sound, even.

“Give me a moment,” I said.

I walked to the small sink in the corner and splashed cold water on my face. No makeup. No filter. I stared at myself in the cracked mirror. Messy ponytail. Dark circles. The small scar above my eyebrow from a childhood fall off a bike. I knew my own metrics by then. I had memorized the scoring system so thoroughly that I calculated my own face without meaning to. It seemed like it should be either funny or sad. I still couldn’t decide which.

I went back to the desk.

“Take the photo,” I said.

Alex raised the phone. The flash was off. The light was bad. Just the desk lamp and the cold glow of two screens. I looked at the camera with the expression that is not quite a smile and not quite neutral. He took the picture. A little too close, but a test is a test.

He sent the photo to my email without comment. I sat down in front of the screen. I created a new test account. Name: Lena Voss. Age: 29. I uploaded the photo and asked Alex to trigger the updated matching system from the admin panel, for my profile only. The laptop fans pushed harder. The screen went white for several seconds.

Results appeared.

Suggested Matches: 3

The top one showed 94.7%.

Name: Mark Reilly, 32. Software engineer based in Palo Alto. His profile photo showed strong features, greenish-gray eyes, and a slightly asymmetrical smile that radiated quiet confidence. I studied the image the way I had trained myself to do. No obvious filters. No heavy editing. Natural lighting. The proportions were right. Cheekbone dominance. Inter-ocular distance. Jawline contour. It aligned with mine in the way the theory predicted. Not on the surface. Underneath.

His bio said: “Looking for someone who sees beyond the surface. Or at least tries.”

I typed a message before I could second-guess it.

“Hi Mark. The new update says we should talk. I say we test it. What do you think?”

I hit send.

Alex placed a brief hand on my shoulder. “Solid score. Highest we’ve seen.”

I said nothing for a few seconds. Then, quietly: “Thanks for staying with me on this.”

He shrugged, offering a small smile. “Where else would I be?”

Outside, the city kept humming through the night. Ambitious founders chasing unicorns. Investors hunting the next big thing. People like us grinding away in tiny rooms. For the first time in a long while, I allowed myself to believe we might actually be onto something.