Short Story: Approved

Keerthana Nalamothula’s story is a roller coaster ride of corporate emotions we go through at work, balancing life delicately with our profession.

Editor’s Pick

This piece is available for free reading this week

The projector hummed softly, the sound just loud enough to be noticed when no one was speaking. A pale rectangle rested on the far wall, slightly skewed. The lights had been dimmed halfway, enough to see faces.

The conference table ran the length of the room, crowded with water bottles branded with different company logos, charging cables knotted together, and laptops open to inboxes no one had checked in the last ten minutes.

Rohit stood beside the screen, remote balanced in his palm. He had angled himself carefully, one step away from the beam.

“So to integrate AI into our IT services,” he said, clicking forward, “we’ll begin with internal automation of basic ticket triaging, response suggestions, escalation prioritisation.”

Boxes slid into place. Arrows converged. The same terms he spoke appeared verbatim on the slide, bolded.

“This will reduce turnaround time,” he went on, his voice settling into its presentation cadence, “and free up the team for higher-value tasks…. Better productivity overall.”

He smiled, just like he had practised for the whole week, and let his eyes travel down the table. Most people nodded. A murmur of assent moved through the room. Someone at the far end leaned forward, squinting at a block of text that had been shrunk to fit.

Rohit clicked again. Then again. Faster now. The final slide appeared, concluding the presentation.

At the front of the table, the project manager sat slightly apart from the others, sleeves rolled to the forearm, pen resting behind his ear. He wasn’t looking at the screen; instead, he observed the room, already counting how long this meeting had taken.

“And by when can we implement this?” he asked.

Rohit didn’t pause. “Immediately,” he said, already bending to shut his laptop. “We’re ready with the first version rollout.”

The projector blinked and froze on the last slide.

A few people exchanged looks that suggested relief. One chair scraped back an inch. Another meeting was cleanly done.

The manager didn’t speak right away, but let out a hum.

He kept his hands folded on the table, eyes moving slowly from face to face, as though checking whether anyone was about to complicate things. The screen turned blue. No one reached for the switch.

A pen clicked. Someone slid their glass closer, then stopped.

Mehak sat two seats from the end, her notebook open on her lap. She hadn’t written anything. The page was blank, the lines sharp. Her pen lay across it, perfectly straight. She rolled it once between her fingers, then tapped it lightly against the edge of the book.

“Can I ask something?” The words echoed.

A few heads turned. Someone who had half-stood eased back into their chair. The room re-collected itself.

Rohit froze mid-motion, one hand still resting on the laptop lid. He looked at her. The smile stayed on his face for a second too long, then slipped into a stern expression.

“Yeah,” he said. “Sure.”

He straightened, adjusting his tie, shifting his weight back toward the screen.

Mehak sat upright. Her voice was low and even.

“For the escalation prioritisation,” she said, “what’s the fallback if the model misclassifies high-impact tickets? Sometimes clients use a different language. The keywords don’t always reflect urgency.”

The office fell silent. Rohit glanced at the frozen slide, as though the answer might still be there, then back at her.

“The accuracy is pretty high,” he said. “We’ll be training it on historical data with similar words and patterns.”

“Yes,” Mehak said, nodding once. “But historical data reflects older patterns. Some of our legacy contracts have exceptions that don’t surface in tickets until late… Clients sometimes downplay urgency.”

A chair creaked as someone leaned back. Another person folded their arms.

Rohit tilted his head slightly, eyes flicking toward the project manager. “Right,” he retorted. “But those are edge cases.”

Mehak turned a page in her notebook. Both sides were blank.

“If the exception is caught late,” she continued, “the escalation delay could harm us. What’s the mitigation plan in that case?”

Rohit’s expression tightened. He glanced at the team lead.

“We can override manually,” the words gritted between his teeth. “There’ll be human supervision in the initial phase.”

“For how long?” Mehak asked. Her tone didn’t change. “Will this ever be supervision-less?”

Before Rohit could answer, the team lead leaned forward.

“AI needs trained data,” she said. “That’s how it improves.”

“I’m just saying—” Mehak interjected.

“We’re at a point where movies can be made using prompts,” someone cut in, half-laughing. “You really think AI can’t detect urgency levels in a task?”

A few people smiled. Someone nodded.

The room loosened. Bodies angled away from Mehak. Attention slid sideways. The question began to wane.

The manager raised a hand.

“Okay,” he interrupted. “That was good work…Let’s pause. We’ll rethink this part, look at the safeguards, and regroup.”

“But the version is ready,” the team lead reminded.

The manager cleared his throat. “We can refine it further.” His gaze settled on Mehak for a beat longer than necessary. “Everyone needs to be confident about this. Let’s meet next week.”

He pushed his chair back. Someone reached for the switch, and the overhead lights flickered on, harsh after the projector’s glow. The slide vanished, and the room returned to itself, table scratches, coffee rings, the faint hum of the AC.

People began gathering their things. Shoes stomped, and chairs scraped. A few murmured complaints surfaced and dissolved.

Mehak closed her notebook with a soft exhale.

“What’s your problem?” The voice wasn’t loud, but it landed before the scrape of the table had finished echoing.

Mehak looked up.

Rohit had stepped away from the podium, closer now, his badge swinging slightly against his chest. Two others, juniors, had drifted in behind him, forming a loose half-circle.

“Do you know how long we’ve been working on this?” Rohit said. His voice stayed even, “Months.”

Mehak scratched the corner of her notebook with her thumbnail, worrying at a frayed edge of the cover. “I know.”

She’d been here before. This wasn’t the first time someone was upset over her decision, and she couldn’t blame them. They all had the same tone, arms crossing, anger written across their faces.

“Then why now?” one of the juniors asked. “Why bring it up in the meeting?”

The room felt smaller with the projector off.

She felt the familiar tightening in her chest, the point where explanation tipped into defence.

“It’s my role,” she said. “To flag potential risks.”

Someone let out a short, dismissive breath. “You’re not flagging. You’re blocking.”

“I’m trying to prevent—”

“Prevent what?” another voice cut in. “Failure? Or accountability?”

The team lead stepped forward, adjusting his watch as he spoke. “This isn’t research, Mehak. At some point, you stop poking holes and let it run.”

A chair scraped behind her as someone stood up, then paused to listen.

“If there’s an issue,” one of them said, “we’ll fix it. That’s how systems work.”

“But how will we even know the issue unless we launch?” another added. “No model is perfect on day one.”

Mehak nodded slowly. She’d heard these sentences so often that they felt borrowed.

“I’m not saying don’t launch,” she said. “I’m saying it can cause potential issues…”

Rohit’s expression loosened, the tension easing from his shoulders. “See,” he said, gently now, “that’s the problem. You always need conditions.”

“You could’ve just said yes,” one of the juniors said, softer this time. “And sent a note later.”

Mehak looked at him. “And if something goes wrong?”

He shrugged, already half-turned toward the door. “Then we learn.”

She waited for someone to say, “You were right.”

Rohit glanced at his watch. “The manager pushed this because of your questions,” he said. “Now it’s on you to justify delays for things that may never even happen.”

“That’s literally risk assessment,” she said. “You’re asking me to compromise.”

His smile weakened. “Risk assessment is fine. But there’s a way to do it without derailing momentum. We’re working very hard. Trust us.”

“So what would you prefer?” she asked quietly. “I say yes?”

He didn’t answer immediately. Then, “Be supportive. We’re all on the same side.”

The others nodded, relief passing through them like a signal. The pressure simmered. Apparently, the issue was resolved. As they left, someone clapped Rohit on the shoulder. “Good call. We can launch it soon.”

Mehak opened her bag and slid her laptop out, the zipper loud in the now-emptying room. She opened a new document and began to type the risk assessment report.

“will likely fail.”

She paused. Backspaced.

“ May require monitoring.” Backspaced.

“The project can proceed with potential observations.” Sentence by sentence, she softened it. Filed the edges down. Made it safe.

By the time she submitted it, the concern was evident. But it had simply been approved.

About Author