JOB ID:QT-SUL-09-443
Job Title – Senior Fullstack Engineer - Generative AI Solutions Location: Remote (Anywhere in India) Experience: 7+ years Candidate Skills Frontend Craftsmanship • Modern JavaScript/TypeScript expertise - you understand the language, not just the framework. • Deep experience with React (or Vue/Svelte) and state management patterns that don't become spaghetti at scale. • You've built complex, interactive UIs: real-time updates, streaming content, optimistic updates, and graceful error states. • Understanding of web performance: bundle optimization, lazy loading, Core Web Vitals - and knowing when to actually care about them. • Experience with AI-specific UI patterns: chat interfaces, streaming text responses, inline suggestions, human-in-the-loop workflows. • Accessibility isn't an afterthought - you build interfaces that work for everyone. Backend Proficiency • Strong backend experience in Python, familiarity with C# is a plus - you can own the full request lifecycle. • API design that developers want to use: REST, GraphQL, or WebSockets depending on what the problem actually needs. • Database fluency: PostgreSQL, MongoDB, Redis - and understanding when each is the right tool. • Experience with vector databases (Pinecone, Weaviate, Qdrant, pgvector) for semantic search and RAG implementations. • Message queues and async processing patterns - because not everything should block the user. Generative AI Integration • You understand prompting as an engineering discipline: structure, token efficiency, context window constraints, and output consistency. • Experience integrating LLM APIs: OpenAI, Anthropic, Azure OpenAI, or self-hosted models (Llama, Mistral, Qwen). • You know models aren't interchangeable - structured output support, instruction - following quality, and quirks vary significantly. • Familiarity with streaming responses and how to build UIs that feel responsive while tokens trickle in • Understanding of RAG patterns: document chunking, embedding strategies, retrieval optimization, and reranking • Experience implementing AI guardrails on the frontend and backend: input validation, output filtering, content moderation UIs AI Security & Privacy Awareness • You've thought about prompt injection - and built defenses against it. • Understanding of PII handling in AI contexts: what should never reach an external model, masking strategies, and audit requirements. • Experience building user-facing controls for AI features: consent, transparency, and correction mechanisms. Agent & Conversational AI Understanding • Familiarity with building conversational interfaces that maintain context across turns. • Understanding of tool/function calling patterns and how to surface agent actions in the UI. • Experience with memory patterns: conversation history management, summarization triggers, and long-term context retrieval. • You've seen (or built) multi-agent systems and understand the UX challenges of surfacing complex AI workflows to users. Engineering Standards • Testing is how you ship with confidence: unit, integration, E2E, and you've probably added visual regression testing too. • CI/CD pipelines are part of your workflow - you contribute to them, not just use them. • You can provision infrastructure when needed and understand deployment beyond "push and pray". • Quality is your responsibility - "QA will catch it" isn't in your vocabulary. Capabilities • Design and build end-to-end features for AI-powered products - from database schema to polished UI. • Create responsive, accessible interfaces for AI interactions: chat, suggestions, document processing, human-in-the-loop reviews. • Implement streaming architectures that make AI responses feel instant, even when they're not. • Build prompt management systems with versioning, testing, and performance tracking across the stack. • Develop context window optimization strategies - intelligently selecting what context to include and what to summarize. • Implement client-side and server-side AI guardrails: input sanitization, output validation, PII detection, and user-facing controls. • Design abstraction layers that allow model switching without full rewrites - your code shouldn't be married to one provider. • Build conversational memory systems: context windowing, semantic retrieval, and conversation summarization. • Create UIs that surface agent behavior transparently - users should understand what the AI is doing and why. • Optimize for real-world performance: caching strategies, request deduplication, and graceful degradation when AI services hiccup. • Contribute to design systems and shared components that accelerate AI feature development across teams. Responsibilities • Own features end-to-end: from understanding the business need through implementation, testing, deployment, and iteration. • Build and maintain frontend experiences for our generative AI products - with obsessive attention to responsiveness and usability. • Develop backend services that integrate LLMs reliably: error handling, retries, fallbacks, cost tracking, and observability. • Implement AI security patterns: prompt injection prevention, sensitive data masking, output filtering, and compliance logging. • Create and maintain prompt libraries with proper versioning, A/B testing infrastructure, and regression detection. • Design APIs that abstract AI complexity appropriately - giving consumers power without exposing implementation chaos. • Collaborate with AI Engineers on model capabilities, limitations, and production deployment patterns. • Build monitoring and analytics for AI features: usage patterns, quality metrics, cost attribution, and user feedback loops. • Maintain comprehensive test coverage: unit, integration, E2E, and AI-specific evaluations (output quality, prompt regression). • Participate in architecture decisions and shape our evolving AI development practices. • Stay current with the rapidly moving AI + web development landscape - and share what you learn. • Embrace the pace: features pivot, priorities shift, and yesterday's work might get shelved - you're energized by this, not defeated.