Insights

Perspectives on data, AI and agents, industry by industry.

79 points of view from our industry teams — filter by industry or topic, or search. Figures in each article are sourced or labelled.

Industry

Topic

5 insights · Clear filters

Airlines & travel Point of view · AI in airlines7 min read2026-09-18

Fare rules are the hidden tax on airline retailing. GenAI can finally read them

Every carrier writes its penalties and conditions differently. Rule-based parsers break on each new format, so agencies quote penalties wrongly and disputes follow. Large language models change the economics — if they are validated, not trusted blindly.

99.8%fare-rule accuracy with our AI engine, against an industry average of 82%
Airlines & travel Point of view · Airline industry trends6 min read2026-09-04

Virtual interlining: selling the itineraries the network can't

Budget travellers will take more stops for a lower fare, but code-share itineraries price them out. Treating the global flight network as a graph — with real connection rules built in — opens itineraries traditional channels never show.

392virtual route combinations from 22 airlines in an illustrative demo search
Airlines & travel Point of view · AI in airlines6 min read2026-08-21

Disruption recovery with supervised agents: faster options, same accountable people

When the operation breaks, recovery is still pieced together by phone and spreadsheet. Agents can assemble ranked options and rebook passengers within policy — but the duty manager must stay the one who decides.

Airlines & travel Point of view · Transformation5 min read2026-07-30

Revenue accounting without the month-end scramble

Tickets, coupons, EMDs, settlements and interline billing are still reconciled across many systems. Matching, integrity checks and commentary are where agents pay back first — with controllers approving every entry.

Airlines & travel Point of view · Future of airlines5 min read2026-07-09

Governing AI in a safety-critical business

Airlines can't let an opaque model decide a maintenance finding, a crew assignment or a passenger's refund. The answer is not to avoid AI but to give every agent an autonomy level, an owner and an audit trail.