Make AI Do the Actual Work, Not Just Talk About It

Own the entire AI inference stack from user message to negotiation to delighting the user by getting the job done.

This means :

- When users in rural Rajasthan asks about complex credit decisions at 11 PM, your code determines if they get an answer in 500ms or 5 seconds.

- When our AI needs to negotiate with banks, lenders and credit bureaus simultaneously for 10,000 users, your architecture decides if we succeed.

- You'll be scaling it the way you want the agent the respond to your friends and family.

Must-Have Skills :

- 3+ years React Native with production apps shipped to both app stores.

- Strong WebSocket integration experience - connection lifecycle, reconnection strategies, and event-driven message handling on mobile.

- Local data persistence using SQLite or WatermelonDB, including schema migrations and offline-first patterns.

- TypeScript proficiency - defining strict types from backend API contracts and maintaining type safety across the data layer.

- Performance optimization on Android - FlatList tuning, memory profiling, and Hermes engine familiarity.

- State management with Redux or equivalent - handling concurrent async state (network, WebSocket, local DB) without race conditions.

- Experience rendering polymorphic UI components - mapping different message types (text, rich cards, suggestions, documents) to distinct components from a single data stream.

Key Growth Areas :

- Implement the chat interface for following established architecture patterns and design specs from our UI/UX team.

- Design the real-time communication layer - a transport-agnostic message pipeline that works across WebSocket, HTTP polling, and future push notification channels without UI-layer changes.

- Architect the local data layer - offline-first storage, sync strategies, conflict resolution, and schema migration patterns that scale as Dhruva adds new capabilities.

- Build extensible rendering infrastructure for AI-native features - streaming responses, polymorphic message types, inline actionable widgets (payment CTAs, score visualizations), and proactive nudge surfaces.

- Handle mobile-specific edge cases - app backgrounding/foregrounding, token refresh, graceful degradation on network loss, and error recovery UX.

- Write well-tested, documented code that the team can maintain and extend as Dhruva's capabilities grow.

Interview Process :

1. Technical Assessment :

- Deep-dive on past work (60 minutes).

2. System Design :

- Architecture and scaling conversations (60 minutes, preferred : in-person).

3. Final Round :

- Cultural alignment and team interaction.