The commute is the product—routing as the quiet operating system of corporate India.

Walk into the transport desk of a Bangalore tech park and the planning happens in Excel — driver names in column A, vehicle numbers in column B, routes in C through M, repeated for the morning, midshift, and night runs. WhatsApp groups carry the live overrides. The system works in the sense that buses leave on time. It also leaves ten to thirty percent of seats empty because no human can solve a vehicle-routing problem with eight hundred origin–destination pairs and six constraints inside a working memory.

Routematic's first sales conversation is not against another vendor. It is against the inertia of "we already do this." Every demo therefore needs to surface the gap in seat utilisation before anything else — show the spreadsheet that the customer already runs and prove it is leaving money on the table.

The naive framing of employee transport is "pick the shortest path." The actual problem is closer to a vehicle-routing problem with time windows, capacities, gender rules, and shift-start guarantees. A solver has to assemble routes such that every employee is picked up inside a fifteen-minute window, no woman is the first pickup or last drop after a certain hour, no detour exceeds a defined ratio, and the vehicle reaches the campus before the 9:00 a.m. badge-in cutoff. Each constraint roughly doubles the search space.

Routematic's wedge is not raw distance optimisation; the open-source solvers do that competently. It is constraint compliance at scale — proving, for each shift, that every route obeys every rule. That proof is the artefact the auditor signs off on, and the auditor is the actual buyer.

The post-2013 corporate rule in India for women employees on late shifts — that a female employee cannot be the first pickup or the last drop, that a marshal must travel after a certain hour, that real-time GPS and panic buttons must be auditable — is the single most cited constraint in the conversation. Most legacy systems treat it as a post-hoc filter: build the route, then check the rule, then patch the violations. Routematic's posture is to build the rule into the solver itself, so a violating route is never generated.

The difference is operational. A filtered system produces ten percent route reruns and a constant trickle of incidents requiring HR review. A constrained-solver system produces neither. The buyer rarely articulates this preference but feels it the first time a shift goes a quarter without an incident escalation.

Routematic does not own a single vehicle. The cabs and buses belong to fleet operators — small businesses with five to fifty vehicles each, scattered across every Indian metro. Routematic onboards them onto its platform, runs background checks on drivers, certifies vehicles, and routes them as a virtual fleet. The operator carries the depreciation; the platform carries the decision.

This is the textbook intelligence-above-the-asset move. A fleet-owning competitor — and several have tried — faces the same financing cost on every new city and slows down as scale rises. The aggregator slows down only on integrations, which compound. By city seven the math diverges in a way the fleet-owner cannot recover from without abandoning the model.

India added an estimated 1.6 million global capability centre employees in the last five years and is on track for over two million by 2030. Each new GCC needs a transport partner from week one — before it has its first thousand employees, before it has its second floor, before HR has hired the people who will eventually run procurement. The buying window is short and the alternative is the spreadsheet.

Routematic's go-to-market is calibrated to this rhythm: a small specialist team that follows commercial leasing signals into new campuses and signs the transport contract before the building is occupied. The growth rate of the category is not Routematic's growth rate. It is closer to the floor of it.

Three years ago, the RFP for an employee-transport contract had a price line and a service-level line. Now it has a carbon line. Multinational buyers report Scope 3 emissions to their group sustainability teams; Indian listed companies disclose BRSR data; both want a per-seat carbon number on the contract. The number used to be unknown; now it is a tiebreaker.

Routematic's edge is that the platform already has the data — distance per route, vehicle type, occupancy, fuel — and can serve a defensible per-seat carbon figure as a report rather than an estimate. Competitors who started with spreadsheets cannot produce the same number without a six-month consulting engagement. The tiebreaker becomes a moat by accident.

Every conversation about employee-transport EV ends in the same place: a pilot of fifty to a hundred vehicles, restricted to short urban routes, charged at depots the fleet operator owns. Range anxiety on a Bangalore-to-Whitefield-via-ORR run is real; battery degradation across two shifts a day is real; the per-kilometre economics are favourable until you price the charging infrastructure. Routematic's role is to schedule around all of it — assigning EVs to the routes whose distance and dwell time fit the battery and the diesel cabs to everything else.

The platform position here is quiet but important. A transport buyer cannot operate a mixed fleet without software that decides which vehicle goes where. The EV transition does not displace the routing layer; it makes the routing layer load-bearing.

The 2020–2022 thesis from every consultancy was that hybrid work would shrink the employee-transport market by half. The reality is messier. Total seat-days fell. But the variance in demand exploded: Tuesday and Wednesday now run at peak, Friday at thirty percent, with employees self-selecting through HR booking tools. A fixed fleet sized for the old peak is now wrong every day. A solver that can re-route around a daily-shifting demand pattern is suddenly worth more, not less.

Routematic's product roadmap pivoted to variable-demand scheduling roughly when the consultancies were writing the obituary. The buyers who survived hybrid did not want fewer routes; they wanted routes that knew which days were Tuesday.

HR signs the contract. The CFO reads the line. Historically the conversation between the two was thin because employee transport sat inside "facilities" — a fuzzy line with no per-head visibility. Routematic's dashboard surfaces per-head, per-route, per-shift cost, which lets the CFO start asking questions HR had not been asked before: why does Bangalore South cost forty percent more per seat than Bangalore North; why does Friday utilisation suggest closing two routes; why is the overnight shift's per-seat number what it is.

The product effect is interesting. The CFO becomes a second buyer inside the account. Renewals stop being an HR procurement formality and start being a finance review. That changes the win rate of competitors who only sell to HR — they do not have the numbers the finance review now expects.

Transport is a state subject in India. Karnataka's permits do not work in Telangana. Maharashtra has its own rules on commercial passenger vehicles. Tamil Nadu has a different stance on app-based aggregator licences. Each new city is a regulatory diligence project — not a deployment. Routematic's scaling cost is therefore lumpy: the first hundred customers in Bangalore are cheap, the first one in Hyderabad costs three months of compliance work, the same in Pune costs another two.

This is where the aggregator model pays again. Local operators already hold the local permits; the platform's job is to verify, surface, and route around the local rules. The operator's existing licence is, effectively, the platform's market-entry document.

The most fragile node in employee transport is the driver. A cab operator can survive a missed week. A driver cannot. When fuel rises, when a shift gets cancelled, when payments are delayed, drivers leave for the next-best wage — often a passenger aggregator paying the same per hour without the night discipline. The platform that wants to scale has to keep the driver pool warm in the soft quarter that always comes.

The mechanics are unglamorous: faster payment cycles to operators, audit-grade trip logs that defend the operator's invoice, on-time settlement during festive months. The deeper insight is that the platform's competitive position is set at the driver layer, not the enterprise layer. Lose drivers in a soft quarter and the next RFP comes with a service-level you cannot honour.

A large GCC's internal audit team reviews employee transport quarterly. They want to see: every driver's verification on file; every vehicle's commercial permit valid; every female employee's late-shift route compliant with the company's safety policy; every incident logged with response time. The platform that produces these reports in a click wins the audit. The one that requires a three-week extraction loses the renewal regardless of how good the routing is.

This is why the transcript-level work — every trip, every detour, every panic-button press logged with timestamps — is the part of the product that does not show in a demo but decides every contract above a certain size. The user-facing app is theatre. The audit export is the contract.

The classical employee-transport unit is a thirty-seater bus from a nodal point to the campus. That model works when employees cluster. In a Bangalore where a single GCC draws from forty pin codes, clustering is harder. The new surface is first-mile: a smaller vehicle picks the employee up from home and drops them at the nodal point; the bus does the bulk haul. The platform schedules both legs as one journey.

The interesting economic effect is that first-mile vehicles can be smaller, electric, and operated by even smaller local operators. The platform's supply pool widens; the customer's per-seat cost stays flat because the load factor on the bus rises. Both directions of the trade improve at once — a rare structural win.

Day shifts are forgiving. Buses run hot, employees flex, a fifteen-minute delay is a Slack message, not an incident. The night shift is the opposite — a missed pickup at 1:00 a.m. in Whitefield is a safety event, a delayed drop at 5:00 a.m. in Hyderabad's HITEC City means an employee waits alone outside a closed gate. Every constraint in the system tightens. Every margin for error contracts.

Routematic's posture on the night shift is that the routing solver has to plan for it as the default difficulty, not the edge case. The product feature that makes the day shift twenty percent cheaper to operate is interesting; the one that makes the night shift safe is buy-or-die. The night shift, in other words, is where renewal decisions are actually made.

Sitting at the centre of a routing platform across forty campuses gives a particular kind of data: the actual commute time between any two points in a city, on any given hour, averaged across millions of trips. That is more current than what TomTom or Google Maps surfaces for the same corridor because it reflects shuttle-class vehicles, not passenger cars, and it is collected on the exact schedule when traffic load matters.

The category-adjacent move is to expose that corpus — anonymised, aggregated — as an input for city-planning teams, real-estate developers, or transit authorities. The product surface is not yet built, but the data substrate exists. The platform that gets to it first picks up a position no fleet operator can replicate, regardless of vehicle count.