c. 30 YEARS EARLIER. SINGAPORE.

  The anomaly appears for less than half a second.

  Lakshmi almost misses it.

  The fetal monitor scrolls smoothly - heart rate, oxygen saturation, uterine pressure - its soft green waveform reassuring in its regularity. A good reading. A boring one. Exactly what you want at seven-thirty in the morning on a Wednesday when the audit team is due at nine and the coffee machine on Level Four is still broken.

  Then, just before the line refreshes, a narrow spike rises where it shouldn’t.

  Not high enough to trigger an alert. Not long enough to register as an event. A deviation so brief it would vanish into noise if you weren’t watching for it.

  Lakshmi leans closer to the screen.

  The spike is already gone. The waveform has smoothed itself, self-corrected, as if embarrassed to have drawn attention.

  “Did you see that?” she asks.

  The technician beside her - young, earnest, on her second rotation - blinks at the display. “See what?”

  Lakshmi rewinds the buffer. Slows the playback. Points.

  “There,” she says.

  The technician squints. “That? That’s within tolerance. Probably a sensor artifact. Happens when the mother shifts.”

  Lakshmi nods. Sensor artifacts do happen. Especially on older equipment, and especially in a wing that’s been promised upgrades for three consecutive budget cycles without ever receiving them.

  Still.

  She watches the loop again. The spike doesn’t look jagged. It doesn’t look like interference. It looks… intentional. Not in the mystical sense - Lakshmi has no patience for that - but in the statistical one. As if the system had briefly sampled a deeper resolution, then decided it wasn’t necessary.

  “What cohort is this?” she asks.

  “Group C. Low-risk. Control-adjacent.”

  Control-adjacent. The phrase makes her frown. There shouldn’t be anything “adjacent” about a control group.

  She straightens, glancing around the monitoring room. The morning shift is settling in: low voices, rustle of paper, the distant hum of air conditioning struggling against Singapore’s humidity. Everything smells faintly of disinfectant and stale kopi.

  “All right,” she says. “Log it. Mark it as transient variance.”

  The technician hesitates. “Do you want me to flag it?”

  Lakshmi considers. Flags create questions. Questions create delays. Delays, on days like this, create scrutiny.

  “No,” she says finally. “Just note it. If it repeats, we escalate.”

  The technician nods and types.

  Lakshmi watches the monitor for another minute. The waveform behaves itself, bland and compliant. She exhales slowly and moves on.

  She tells herself she’s satisfied.

  She is not.

  By eight-thirty, the lab is fully awake.

  Lakshmi moves through it with practiced efficiency, her ID badge swinging lightly against her chest. She knows this place intimately - the scuffed linoleum near the freezers, the way the overhead lights flicker on the left side of Corridor B, the precise angle at which you have to turn the handle on the autoclave to keep it from sticking.

  She helped design three of the protocols currently running here. Reviewed another dozen. Fought for stricter consent language when the administrators wanted something leaner, more “participant-friendly.” Won that argument, too.

  This matters. She is not a junior researcher dazzled by access or funding. She is not careless. She believes in process. In guardrails. In the slow, unglamorous work of making sure medicine does more good than harm.

  That belief has carried her through fifteen years of biotech, from grad school to government-adjacent research, from optimism to something harder and more durable.

  Which is why the unease from the monitor refuses to leave her.

  She stops at the glass wall overlooking the main analysis floor. Rows of desks, terminals glowing softly, researchers bent over screens. No chaos. No drama. Just work.

  This project - officially titled the Maternal–Fetal Outcome Stabilization Trial - is supposed to be one of the cleanest they’ve ever run. Low intervention. High oversight. Conservative endpoints. The kind of study you cite in ethics lectures as an example of how to do things properly.

  It is also ending today.

  By this afternoon, the data will be locked, the equipment decommissioned, the remaining biological samples either destroyed or transferred according to the retention agreement.

  Closure. Lakshmi likes closure.

  She turns away from the glass and nearly collides with Sumitra.

  “Lakshmi,” Sumitra says, steadying herself with a laugh. “I should have known you’d be haunting the monitors.”

  “Someone has to,” Lakshmi replies. “You’re early.”

  Sumitra shrugs, pushing a loose strand of hair back into her bun. “Couldn’t sleep. The audit’s making everyone twitchy.”

  “Are you ready for it?”

  Sumitra smiles, confident and unforced. “Of course. The data’s solid. Consent is airtight. If they find anything, it’ll be procedural trivia.”

  Lakshmi studies her colleague’s face. Sumitra looks tired, but energized in the way people do when they’re certain of their footing. She has always been like this - brilliant, meticulous, fundamentally optimistic about the ability of good science to justify itself.

  If you discover this tale on Amazon, be aware that it has been stolen. Please report the violation.

  “I saw a variance spike this morning,” Lakshmi says.

  Sumitra’s smile barely flickers. “Which cohort?”

  “Group C.”

  “That’s well within noise margins,” Sumitra says immediately. “We’ve seen micro-fluctuations there before.”

  “Have we?” Lakshmi asks.

  Sumitra hesitates for a fraction of a second. “Statistically.”

  Lakshmi lets it go. This is not the hill to die on at eight-forty-five with auditors inbound.

  “Elias is already here,” Sumitra adds, as if anticipating the next question. “He’s in Conference Room Two with the liaison.”

  Of course he is.

  Lakshmi thanks her and continues down the corridor, her unease now braided with something sharper.

  Elias Crowe does not look like someone who should make her uneasy.

  He is polite. Soft-spoken. Meticulous in his dress in a way that suggests habit rather than vanity. He never raises his voice. Never interrupts. When he disagrees, he does so by asking questions so precise they feel less like challenges and more like inevitabilities.

  He is, officially, a consultant - external systems theorist brought in to advise on data modeling and long-term outcome interpretation. He does not have decision-making authority over the trial. He does not sign off on protocols.

  And yet.

  When Lakshmi enters Conference Room Two, Elias glances up from the table and inclines his head slightly, as if he had been expecting her.

  “Dr. Menon,” he says. “Good morning.”

  “Dr. Crowe,” she replies. “I didn’t realize you’d be joining us for the closeout.”

  “I thought it prudent,” Elias says mildly. “Given the retention questions.”

  The liaison - a man in a crisp shirt who looks perpetually apologetic - clears his throat. “Yes. We were just reviewing the archival provisions.”

  Lakshmi takes a seat, folding her hands on the table. “The retention terms are standard,” she says. “Anonymized samples may be stored for longitudinal analysis, subject to oversight.”

  “Yes, but - ” the liaison begins.

  Elias interjects gently. “The concern, as I understand it, is scope. Whether the retained data exceeds what is strictly necessary for the trial’s stated outcomes.”

  Lakshmi meets his gaze. “It does not.”

  Elias tilts his head. “Perhaps not for the primary endpoints. But for secondary analysis - false negatives, latent correlations - it may prove invaluable.”

  The liaison looks relieved to have someone else carrying the argument. “Exactly. We just want to be sure everything is justified.”

  Lakshmi’s jaw tightens. “We already justified the retention in the ethics submission.”

  “And it was approved,” Elias agrees. “Unanimously.”

  There is something in the way he says it that makes Lakshmi pause. Not triumph. Not manipulation. Simply… certainty.

  “What exactly do you intend to analyze longitudinally?” she asks.

  Elias does not hesitate. “Inheritance stability. Marker persistence across non-expressive generations.”

  The words land softly, but they land.

  “That’s outside the trial’s scope,” Lakshmi says.

  “Formally,” Elias agrees. “Which is why the data would remain anonymized and non-interventional. Observational only.”

  The liaison nods vigorously. “Purely academic interest.”

  Lakshmi feels the familiar pressure building behind her eyes. “Observational studies still require purpose.”

  “They require possibility,” Elias corrects. “Purpose emerges later.”

  She looks at him sharply. “That’s not how ethics works.”

  Elias smiles faintly. “No. But it is how knowledge does.”

  The room falls quiet.

  Lakshmi is acutely aware of the clock on the wall. Of the auditors waiting downstairs. Of the reality that this argument, however important it feels, is not going to be resolved here.

  “Fine,” she says at last. “The retention stays as approved. But any secondary analysis requires renewed ethics review.”

  “Of course,” Elias says, without missing a beat.

  She does not like how easy that was.

  By late morning, the audit is underway.

  Lakshmi fields questions with practiced calm, flipping through binders, pulling up records, explaining methodologies she could recite in her sleep. Everything checks out. Consent forms signed and countersigned. Intervention thresholds conservative to the point of dullness. Adverse events logged and resolved.

  If there is something wrong here, it is not something the auditors are going to find.

  During a lull, Lakshmi slips back into the monitoring room. The technician from earlier looks up.

  “No repeats,” she says quickly. “Everything’s been clean since.”

  Lakshmi nods, but her eyes go to the screen anyway. She watches for several minutes, looking not for spikes now, but for patterns. For the absence of patterns. For the way the system seems almost… bored.

  “Has Group C always had this many parameters logged?” she asks.

  The technician frowns. “I think so. Why?”

  Lakshmi points to the sidebar. “These. Resolution tags. They’re deeper than Group A’s.”

  The technician glances. “That’s just the redundancy layer.”

  “For what purpose?”

  The technician hesitates. “Dr. Crowe suggested it. Said it might help catch edge cases.”

  Lakshmi closes her eyes briefly.

  “Did he say which edge cases?”

  “No,” the technician admits. “Just… edge cases.”

  Lakshmi straightens. “All right. Thank you.”

  She leaves before the technician can ask more questions.

  Lunch is taken at a hawker centre two blocks away, the kind that survives on inertia and foot traffic rather than charm.

  Lakshmi and Sumitra sit at a plastic table, bowls of laksa steaming between them. The air is thick with chili and fried garlic and the low roar of lunchtime crowds.

  “You’re quiet,” Sumitra says, slurping noodles.

  “I’m thinking,” Lakshmi replies.

  “That’s rarely good news.”

  Lakshmi snorts despite herself. “The retention terms bother me.”

  Sumitra shrugs. “They always do. That’s your job.”

  “Yes, but - ” Lakshmi trails off, searching for words that won’t sound hysterical. “Do you ever feel like we’re measuring something we don’t intend to use?”

  Sumitra considers. “That’s most research.”

  “No,” Lakshmi says. “I mean… something we don’t intend to acknowledge.”

  Sumitra sets her chopsticks down. “Lakshmi. This is one of the cleanest trials we’ve ever run. If there were anything ethically questionable - anything at all - I would have flagged it.”

  “I know,” Lakshmi says softly. “I trust you.”

  Sumitra smiles. “Good. Then eat your laksa before it congeals.”

  Lakshmi obeys, but the taste is dulled by unease.

  The afternoon passes in a blur of signatures and confirmations.

  By four-thirty, the auditors are gone, satisfied. The liaison thanks everyone profusely and promises follow-up emails that will never be read. The lab begins the slow process of shutting down.

  Lakshmi stands in the data room as the servers wind down one by one, their hum dropping in pitch. Screens go dark. Indicator lights fade.

  She watches as datasets are archived, encrypted, transferred to cold storage. Names stripped. Identifiers replaced with strings of numbers that mean nothing to anyone without the key.

  Except, she realizes, to whoever designed the system.

  “Dr. Menon,” a junior researcher calls. “Can you sign off on the final inventory?”

  Lakshmi takes the clipboard and scans the list. Equipment. Samples. Documentation.

  Her pen pauses.

  “What is this subset?” she asks, pointing.

  The researcher peers over her shoulder. “Those are the retained biologicals. As per protocol.”

  “They’re tagged differently,” Lakshmi notes.

  “Yes,” the researcher says. “Long-term storage. Inheritable markers.”

  “Inheritable,” Lakshmi repeats.

  The researcher nods, oblivious. “Dr. Crowe said - ”

  “Thank you,” Lakshmi says, handing the clipboard back. “That’s all.”

  She waits until the room empties, then turns toward the freezers at the back.

  They stand in a neat row, stainless steel doors humming softly. Each labeled. Each locked.

  She opens the inventory terminal and pulls up the detailed listing. Most entries are scheduled for destruction within six months. A small subset - far smaller than she expected - is flagged for indefinite retention.

  No names. No cohorts. Just codes.

  She scrolls.

  The codes are not random.

  She cannot articulate how she knows this, only that the distribution feels… intentional. Curated.

  Her pulse quickens.

  Lakshmi straightens, the weight of the moment pressing down on her. This is where she should escalate. Where she should call someone. Demand clarification. Freeze the process until questions are answered.

  She thinks of the audit. Of Sumitra’s confidence. Of Elias’s calm assurances.

  She thinks of the mothers who trusted them. Of the children who will never know this room exists.

  There is no smoking gun here. No violation she can point to without sounding paranoid.

  So she does what she has been trained to do.

  She logs the inventory as complete.

  The system accepts the entry without comment.

  As she turns off the lights, the freezers continue to hum, steady and patient.

  Lakshmi pauses at the door, listening to that hum, and feels - just for a moment - the faintest chill run down her spine.

  Then she leaves, locking the door behind her.

  The archive remains.

  She signed the inventory and let the freezers hum.

  Because there was no violation she could name without sounding unhinged.

  Because the trial was “one of the cleanest.”

  Because good people like Sumitra believed it.

  And because Elias had already made sure every question had a reasonable answer ready.

  Now a small, carefully chosen subset of biological samples - tagged for “inheritable markers” - is moving to indefinite cold storage.

  The boundary isn’t terminating civilizations anymore.

  It’s archiving them.

  Questions I’m asking while staring at perfectly normal medical records:

  Lakshmi felt the chill but locked the door anyway. How long until that chill becomes the same gentle northward pull?

  Elias wants “marker persistence across non-expressive generations.” Is the system no longer satisfied with clean endings - is it preparing clean beginnings?

  And the deepest one: if the trial data was always too clean to be real, how many other “successful” medical studies have quietly archived the next generation of marked humans?

  The archived markers wait, patient and dormant.

  And somewhere, three different watchers realize the boundary has started planning for inheritance.

  Stay ethical. Stay vigilant. Stay unmarked - while you still can.

  The author who just double-checked every consent form I’ve ever signed