Commercial aviation · India and the Gulf
Where should Indian carriers deploy their next 100 long-haul aircraft, and can the India-Gulf corridor absorb them?
Compete with the Gulf hubs. Do not fly more aircraft to them. Europe first, North America second.
The corridor stays the prize: of 39.7M Gulf passengers, roughly 8.5M are travelling onward to Europe and North America on a Gulf carrier, a contested pool worth INR 28,900 to 56,700 crore. That traffic is won by flying past the Gulf, not to it.
Every figure is computed in-repo from DGCA, Eurostat and World Bank data, or gated against a primary filing: a status gate refuses any hand-entered number that has not been verified, and it has blocked real work rather than been relaxed. DGCA and Eurostat, measuring the same routes from opposite ends, agree to 2.6%. Ten changes of mind are published with the commit each happened in, including a claim that was on the page and had to be withdrawn, and a reason this project gave itself for not checking its most important number.
india-widebody-window.vercel.app
The same project, as evidence
A self-directed case in the style of BCN Advanced Manufacturing & Services commercial aviation work. Python analysis layer, static site, no PowerPoint and no Excel anywhere in the pipeline.
| From the posting | Where it is evidenced |
|---|---|
| Go-to-market strategy | A named client, five costed options each with what would have to be true, a phased roadmap and a nine-row risk register |
| Industry analysis and competitive benchmarking | India's international market by corridor, carrier and stage length; Indian carriers against Gulf carriers on share, load factor and sector length |
| Model building | Capacity absorption in ASK, corridor breakeven, market sizing to a band across three methods, and a sensitivity on every modelled knob. Reimagined in Python; the repo forbids Excel as output or intermediate |
| Secondary research | Every field sourced, dated and reliability graded in a data dictionary; conflicts flagged rather than resolved to whichever was convenient |
| Judgement where data is thin | An assumption gate, a disputed route quarantined, an unverifiable origin-destination share bounded rather than asserted |
| Storylining | SCQA, answer-first structure, and action titles on every chart enforced by a test |
A script maps the real posting to artifacts and checks each exists. It reports 82%, not 100%. Three requirements cannot be honestly evidenced by one person working alone, and are listed rather than papered over: survey analysis, mentoring analysts, and first-level team management. A fielded-ready conjoint instrument for the softest number in the case is designed and included, and survey analysis is still counted as a gap, because designing a survey is not analysing one.
A bug once misfiled 5.0M passengers a year, a fifth of the entire Gulf flow, into the wrong bucket. All 72 tests passed throughout, because a wrong bucket is still a valid bucket. It was found by chasing an unrelated number. Four of the first five documented corrections came from cross-checking one source against another; none came from the test suite. The eighth was found by counting one surface against another by hand, because the two had no shared key for a test to compare. That is why the reconciliation work is an exhibit on the site rather than an appendix.
india-widebody-window.vercel.app · source and full write-up on GitHub