FirstAuthor ₹999  ₹499 early bird Reserve seat

Live on Zoom Sat 24 & Sun 25 October 2026 7–9 PM IST

Your first paper, from blank page to someday submitted.

Two live days on finding a gap that is actually open, designing an experiment that answers it, writing the paper, and surviving peer review — taught with one real paper's draft, reviewer scores, rebuttal and accepted version on screen.

4 hours live Recording for 90 days Workbook + templates 60 seats

--days --hrs --min --sec
until Day 1

The competition

Nobody taught you to write one. Everybody is writing one.

Coursework teaches you to solve problems. It does not teach you to find one worth solving, or to argue for it in eight pages against reviewers with four other papers due the same night.

32,040

preprints posted to arXiv in a single month — June 2026, a record.

arXiv blog, July 2026

21,575

papers submitted to NeurIPS 2025 alone — 61% more than the year before.

NeurIPS programme chairs, 2025

75%

of them were rejected. 5,290 of 21,575 made it in.

Acceptance rate 24.5%

A good idea is not enough any more. What separates an accepted paper from a desk reject is craft: a gap stated precisely, a claim matched to the evidence, a baseline nobody can object to, and a rebuttal that answers the reviewer instead of arguing with them. That craft is learnable in a weekend. Most people spend three years picking it up by rejection.

Fit

Who this is for — and who should skip it.

Come if you are

  • A final-year UG or PG student with a project you suspect could be a paper
  • A first- or second-year PhD scholar who has not published yet
  • Sitting on results you have never written up
  • Writing a paper now and unsure whether the contribution is real
  • Coming back after a reject and not sure what the reviewers actually meant

Skip it if you are

  • Looking for a guaranteed publication — we teach the craft, not a shortcut
  • Hunting for paid or predatory journals that publish anything for a fee
  • Already published three or more peer-reviewed papers
  • Expecting us to write, or an AI to write, the paper for you
  • Unable to attend live — the exercises are the point

Take-home

You leave with six things you can open on Monday.

Not notes. Working files, filled in with your own research during the sessions.

Sheet

Gap matrix

Your area's papers on one axis, their assumptions and limits on the other. The empty cell is your paper.

Overleaf

Paper template

Venue style file, sections split into their own files, refs.bib, a sample table and figure that already compile.

Prompts

Prompt library

Tested prompts for literature sweeps, related-work drafting and self-review — every one tried on at least two models.

Checklists

Pre-submission review

The self-review pass, a venue cheat sheet, and the red-flag list that catches a desk reject before a reviewer does.

Workbook

Exercise workbook

One exercise per section, plus the Day 1 homework you bring back to Day 2.

Recording

Both days, 90 days

Full recordings and every slide, so you can replay the LaTeX and rebuttal sections while you write.

Curriculum

Two days, in the order the work actually happens.

Day 1 decides what you are writing. Day 2 writes it and sends it.

Saturday 24 October · 7–9 PM IST · bring three papers from your area

1.1

What actually counts as a contribution

Five contribution types reviewers recognise, and the four that look like contributions but are not.

20 min
1.2

Reading a field in an afternoon

Search strategy, citation chaining, and where LLMs help versus where they invent references. Live demo.

30 min
1.3

Building your gap matrix

Hands-on: you fill the sheet with your three papers while we do the same on screen.

30 min
1.4

From gap to hypothesis to experiment

Turning a hole in the literature into a claim, and a claim into the smallest experiment that could falsify it.

25 min
1.5

Baselines, ablations and the questions reviewers always ask

What to compare against so the obvious objection is already answered in the table.

20 min
1.6

Live Q&A and homework

Finish your gap matrix overnight. We review a handful on Day 2.

15 min

The running example

One real paper, opened up — including the reviews that hurt.

Most workshops show you finished papers. Finished papers teach you nothing, because every decision that mattered has been erased. We show the draft, the scores, the rebuttal and the diff.

[paper title] Submitted [venue] Accepted after one round

In this paper we propose a novel framework which achieves good results on several benchmarks and outperforms many existing methods. We show that retrieval-augmented models degrade by 14 points when the retrieved passage contradicts the question—a failure mode three prior benchmarks miss because they never test it. We introduce a 2,400-item diagnostic set and a fine-tuning recipe that recovers 11 of those points.

Struck text: the version that got the low scores. Marked text: the version that got in.

First submission Soundness 2 Contribution 2 Presentation 3 → After rebuttal Soundness 4 Contribution 3 Presentation 4
01

The draft we sent

Vague claim, no diagnostic, baselines chosen for convenience.

02

Three reviews

Read line by line, including the one that misread the method.

03

The rebuttal

What we conceded, what we defended, and the one experiment we ran in four days.

04

Accepted

The diff between the two versions, sentence by sentence.

Who is teaching

Two published authors, not two marketers.

Both of us are PhD scholars who went through this last year. Every paper we cite, you can open.

Balaji Venktesh

Balaji Venktesh

PhD scholar · Small language models for code

First author of Capacity vs. architecture: an evaluation of SLMs for automated docstring generation, in Automated Software Engineering (Springer, 2026), and co-author on a second paper in Information Processing & Management.

Autom. Softw. Eng. Inf. Process. Manage.
Gireesh Sundaram

Gireesh Sundaram

PhD scholar · Agentic AI for software engineering

First author of Epistemic fortitude: architectural specialization for sycophancy mitigation in software engineering agents, in Information Processing & Management (Elsevier, in press), and co-author on a second paper in Automated Software Engineering.

Inf. Process. Manage. Autom. Softw. Eng.

Format

Everything you need to know before you pay.

When
24 & 25 OctSaturday and Sunday, 7–9 PM IST
Where
ZoomLink emailed 23 Oct; waiting room on
Seats
60Capped so live Q&A stays useful
Recording
90 daysPlus slides, templates and workbook

Questions

The things people ask us in DMs.

I have never used LaTeX. Will I keep up?

Yes. We start from an Overleaf project that already compiles and you edit it live. You never install anything.

Does this only work for computer science?

The craft — gap finding, claims, structure, rebuttals — is the same across empirical fields. Our examples come from software engineering and machine learning, and we ask your research area on the registration form so we can pick examples close to the room.

Will you help me publish my specific paper?

Not inside the two days — there is not time to do it justice for everyone. We offer one-on-one sessions separately, introduced at the end of Day 2.

Can I attend only one day?

You can, but Day 2 builds on the gap matrix you fill in on Day 1. The price covers both either way.

What if I miss a session?

Both days are recorded and the link reaches you within 24 hours. You keep access for 90 days.

Is using AI to write papers even allowed?

It depends on the venue, and the rules changed recently. We go through the current disclosure requirements of every venue we mention, and we re-check them the week before the workshop.

Do I get a certificate?

Yes, after both days, sent with the recording and materials.

How do I pay, and is a refund possible?

Cards, UPI, net banking and wallets through Razorpay. Full refund on request before 21 October. Details in the refund policy.

Price

Less than one rejected submission fee.

One price, both days, everything included. No upsell required to follow along.

Early bird — open now

₹499

Until 11 October, 11:59 PM IST

  • Both live days, 4 hours total
  • Recording access for 90 days
  • Workbook, Overleaf template, gap matrix, prompt library
  • Certificate of completion
  • Student WhatsApp group and post-workshop Q&A
Reserve your seat
60 of 60 seats left
After 11 October

₹999

Same workshop, same materials

  • Everything in early bird
  • Subject to remaining seats

Full refund if you cancel before 21 October, or if we cancel the workshop. See the refund policy.

Register

Take a seat.

Pay first, then tell us about your research so we can tune the examples to the room.

1. Pay ₹499

Card, UPI, net banking or wallet. Secured by Razorpay. You get a confirmation email immediately.

Questions before you pay? Write to hello@firstauthor.in or message us on WhatsApp.

2. Tell us about your work

Takes a minute, and it tunes the examples to the room. It also gets you The six-sentence abstract — three pages, free, whether or not you register.

  • The formula: one job per sentence, with the trap each one falls into
  • Two abstracts rewritten from the version that scored 2s — one in machine learning, one in power systems
  • Structured abstracts, keywords, titles, word swaps and the desk-reject tripwires
  • A worksheet to write your own six in, and two prompts that make an LLM critique it instead of writing it