Work setting : Work from office (WFO).

Purpose of the Job :

We are building a new generation of AI-native platforms where AI is not an add-on - AI is the core system. These platforms reason, retrieve, plan, call tools, and act across chat, voice, and video. They operate on streaming data, integrate deeply with product workflows, and must meet strict expectations around latency, correctness, reliability, and privacy.

The foundations are in motion. Now we need builders who can own the next phase - engineers who design systems, not just features; who ship weekly; who think end-to-end; and who take personal responsibility for what runs in production. This is a role where you define what gets built, how it is built, and why it matters. You will work with a lean, high-velocity team using Python, GoLang, Next.js, MongoDB, Redis, Snowflake, streaming data platforms, and vector databases to create multimodal, agentic, privacy-first AI systems.

Key Responsibilities :

1. System & Product Ownership :

- Own large product surfaces and backend domains end-to-end.

- Translate ambiguous problems into concrete, shippable systems.

- Break big ideas into weekly deliverables.

- Define architecture, data models, APIs, and UX flows.

- Own quality, reliability, and operational readiness.

2. AI-Native Engineering :

- Build and optimize RAG pipelines (chunking, embeddings, re-ranking, caching).

- Design agent pipelines (planner, retriever, tool executor, validator).

- Build multimodal AI experiences across chat, voice, and video.

- Tune systems for low latency, high throughput, and predictable cost.

- Treat prompts, schemas, and evaluation datasets as first-class artifacts.

3. Backend, Data & Infrastructure :

- Build low-latency services in Python and GoLang.

- Develop full-stack product surfaces using Next.js.

- Use MongoDB for operational data and Redis for caching, queues, and ephemeral state.

- Build streaming ingestion pipelines for near real-time freshness.

- Use Snowflake for analytical workloads.

- Implement retrieval layers using vector databases such as Milvus and Pinecone.

4. Privacy, Security & Trust :

- Design privacy-first architectures.

- Implement access controls, encryption, and audit logging.

- Ensure AI systems respect data boundaries and consent.

- Build guardrails against data leakage and prompt injection.

5. AI-Assisted Development :

- Effectively utilize AI-enabled tools such as Cursor and Claude Code.

- Use AI collaboration for planning, scaffolding, refactoring, test generation, and exploration.

What We Are Looking For :

1. Must-have :

- 3+ years building production software.

- Proficiency with AI-enabled tools such as Cursor and Claude Code.

- Strong backend experience in Python or GoLang.

- Comfortable with modern frontend (Next.js).

- Experience with MongoDB, Redis, and cloud infrastructure.

- Exposure to LLM-powered systems and RAG pipelines.

- High ownership and builder mindset.

2. Strong differentiators :

- Built 0- 1 products or platforms.

- Experience with multimodal AI (chat, voice, video).

- Experience with streaming data pipelines.

- Demonstrated advanced use of AI coding assistants.

- Experience building low-latency systems.

What You Are Like :

- You default to shipping over planning.

- You thrive in ambiguity.

- You care deeply about craft and reliability.

- You enjoy owning problems end-to-end.