Something we actually checked, at a time we can tell you.
TRUST & METHODOLOGY
A fact is a fact. An estimate is an estimate.
YaatriGan is built around one rule: uncertainty should be visible, not dressed up as certainty.
A prediction stays a prediction, and often stays a range.
Worked out from ten thousand simulations of the whole trip.
Missing evidence never quietly becomes confidence.
Four rules that matter more than a flashy score.
1. Current truth outranks predictions.
If the railway currently says WL 35, then it is WL 35 — no model, however confident, is allowed to overwrite it. A prediction can sit beside a fact and tell you what it might become. It cannot replace it.
2. Probability and preference are different numbers.
Journey Success is a probability. The Recommendation Score is about what you said matters — safest, fastest, cheapest. Telling us your priority changes how options are ranked for you; it never changes what is factually likely to happen.
3. Ranges stay ranges.
When the evidence only supports “somewhere between 60% and 85%”, that is what you see. Turning it into a confident-looking “74.3%” would be inventing precision we do not have, and it is the single easiest way to mislead somebody who is trusting a number.
4. Failure degrades honestly.
When something cannot be checked, we show less and say so. We do not fill the gap with a plausible-looking default and let it be read as a finding.
Four kinds of evidence, never mixed.
Everything shown to you falls into one of four buckets, and it is labelled with the one it belongs to:
- Current — what the railway’s own reservation and fare information says right now, with the time we checked it.
- Historical — how trains on this route have actually been running: typical delays, how demand behaves on dates like yours.
- Predicted — an estimate of something that has not happened yet, such as whether a waitlisted ticket confirms. Always a range when the evidence is a range.
- Simulated — our own calculation: Journey Success, connection survival, and the ranking built on them.
A plan shows you which bucket each figure came from, and when it was checked, so you can tell the difference between something we know and something we think.
“We don’t know” is a valid result.
No reliable evidence beats false precision. An option that leans too heavily on things we could not establish is deliberately kept out of “Safest” — a recommendation built on defaults is not a safe recommendation, it is a confident one.
Where AI actually fits.
We use AI and statistical modelling for the parts that are genuinely questions of judgement under uncertainty: reading how a train has behaved historically, estimating what a waitlist position is likely to do, and simulating a whole journey ten thousand times to see how often it works.
We do not use it to invent facts. Seat counts, fares, quotas and waitlist numbers are reported, not generated — and where we could not establish one, the answer is that we could not establish it.
An independent product.
YaatriGan is a product of NutanTech Labs LLP. It is not Indian Railways, not IRCTC, and not affiliated with or endorsed by either. Nothing here is an official ticketing service or a guarantee of a ticket outcome. Before you pay or travel, confirm the final booking details with the official booking channel.