| name | flyrank-context |
| description | FlyRank-specific background — what the company does, the search problem, and the hand-written flags story that the ML internship work builds on. Load for context when an assignment references FlyRank's product, flags, or the "beat the rule" framing. (Project-specific: delete this folder when reusing the skill library elsewhere.) |
FlyRank context
What FlyRank is. FlyRank builds content as infrastructure: it researches, writes, publishes
content straight into a client's website, then watches search data and optimizes — the full
cycle, run by algorithms and AI rather than people doing it by hand.
The problem this internship works on. Content that gets found in Google search ranks, then
quietly decays: rankings slip, clicks drop, and most teams notice too late. The valuable
decision is: out of thousands of pages, which one should a human fix FIRST? That is a ranking
problem on messy, real search data — exactly where a learned model can beat a fixed rule.
The flags story (why "beat the rule" is the framing). FlyRank's product already flags pages
today — a health score, quick-win tags, needs-attention flags. These are hand-written rules:
if-this-then-that, thresholds chosen by hand. They work, and they run in production. But rules
run out where signals get many, tangled, and shifting — and no trained model has replaced them
yet. The intern capstone lives exactly in that gap.
Two things this means for your work:
- Product flags are outputs, never inputs. health_score and friends encode a decision
someone already made. Use them as a baseline to beat — never as model features (that's a
circular result; see the leakage skill).
- Honest language is house culture. FlyRank publishes findings as observed / measured /
directional / decision-support — never "we predicted Google's algorithm." Your work follows
the same claim discipline.
The data provenance in one line. Google Search Console + Google Analytics land in BigQuery,
sync daily, aggregate into warehouse tables, and ship to interns as a pseudonymized public-safe
release (details in the flyrank-data skill).
How to verify you have the context
You can answer, in one sentence each: what FlyRank sells, what decision the capstone improves,
why product flags can't be features, and which four words describe allowed claim language.
