Chalo — Recon Dashboard & Operator App

Daffodil Software · Feb 2023 – Jan 2024

Chalo runs intercity bus operations across India. Their field operators needed a dashboard for financial reconciliation and an Operator App for live ground operations — both used in low-bandwidth environments where every kilobyte counts.

I was the sole frontend engineer on this project. Owned every technical decision — architecture, performance budget, deployment pipeline, production support, and the cross-platform strategy. No senior FE to defer to and no playbook to copy.

what I shipped

• Recon Dashboard — finance-side internal tool for contract management, expense reconciliation, and operational transparency. Designed and built it from a blank Figma file; designed the Jenkins pipeline that deployed it across all environments.
• Operator App — cross-platform (web, iOS, Android) using React and Capacitor. One codebase, three deliverables. The operations team used it daily on shared, often older devices.
• Live multi-camera streaming via WebRTC integrated into the operator dashboard, so ops teams could monitor field activity in real time without context-switching to a separate tool.
• 28% bundle reduction via code splitting, aggressive tree shaking, and dependency pruning. This wasn't a vanity metric — it was the difference between the app loading or timing out on the connections operators actually had in the field.
• Onboarded and mentored new team members when the project ramped — owned the architectural-decision document and the "here's how we work" runbook.

what stuck with me

• Performance budgets are a real constraint, not a nice-to-have, for software that ships into low-bandwidth regions. The 28% number unblocked an entire region rolling out.
• Capacitor is underrated. Single-codebase native apps make a small team move like a much bigger one — provided you stay disciplined about touching native bits.
• Being the only person on a project means you get to make every decision and you have to defend every decision. Both of those make you sharper, fast.

what I'd do differently

• I'd push back harder on scope earlier. Being sole FE made me agree to ship more in a single quarter than I should have; some of that work would have been a stronger feature with another month and a half-time partner.
• I'd set up an end-to-end test on the deploy pipeline from day one. The 28% bundle work would have been less stressful if a regression on payload size could fail the build automatically.