Staff AI Engineer

About The Team: Pattern is an ecommerce accelerator that helps global brands grow across marketplaces. We're the AI Transformation Team -- a small R&D group building the connected system of technologies that powers data collection, intelligence, creative automation, and brand strategy across Pattern's operations. We design, build, and operate an ecosystem of interconnected systems that form a closed loop: collecting raw web data, normalizing it into structured intelligence, analyzing marketplace behavior, and activating insights through creative automation and brand strategy. Pattern's rich data assets -- marketplace analytics, consumer behavior signals, and product intelligence across thousands of brands -- flow directly into this ecosystem. We move fast, ship to production, and give engineers uncommon autonomy over hard problems. AI is not a feature we bolt on -- it is the core of what we build. Every engineer on this team works daily with LLMs, embeddings, agentic workflows, and generative pipelines -- and builds them into scalable, production-grade systems. We share what works, challenge each other technically, and raise the bar together. The Role We're looking for a Staff Software Engineer to own critical systems and shape the technical direction of our intelligence ecosystem. You'll go deep on one or two systems -- designing their internals, scaling their infrastructure, hardening their reliability -- while driving architectural decisions that ripple across the full platform. This is a hands-on role with broad influence. You'll ship code daily -- often in partnership with AI coding tools like Claude Code -- driving them to produce robust, test-driven code at a pace that multiplies your impact. You'll set technical standards through reviews and design docs, and collaborate closely with engineers across the team. We expect you to mentor, uplift, and make the people around you better. You'll operate at the intersection of three pillars: deep business understanding and customer obsession, leveraging generative AI to solve real problems, and designing architecture that is robust and scalable. Your decisions here directly shape products used by global brands. This is a hybrid role based in Pune. Competitive compensation and benefits Systems You'll Shape You won't be limited to a single domain. Our ecosystem spans multiple interconnected systems -- here is where a Staff engineer makes an impact: Agentic Data Collection & Competitive Intelligence -- Know everything about a brand's digital presence before a human asks. Build intelligent agents that discover, collect, and keep product and competitor data fresh across marketplaces at scale. Product Intelligence & Attribute Normalization -- Make every product across Pattern's ecosystem describable, searchable, and comparable. Build the semantic engine that turns messy, unstructured product data into a consistent taxonomy using LLMs, vector search, and hierarchical classification. Marketplace Behavioral Analytics -- Answer the question: what makes a customer buy one product over another? Combine consumer behavior data, competitive analysis, and attention modeling to surface the attributes that actually drive conversion. Creative Automation & Generative Pipelines -- Turn structured product intelligence into marketplace-ready content at scale. Build the systems that connect data to AI-generated imagery, video, and copy -- from model training to quality assurance to final output. Cross-System Architecture -- Make the ecosystem greater than the sum of its parts. Design the shared contracts, communication patterns, and feedback loops that keep these systems connected, observable, and self-improving. Our Stack We hire engineering judgment, not keyword matches. If you've built serious systems on a different stack, we want to talk. Languages & Frameworks: Python, FastAPI, SQLAlchemy, React/Next.js, TypeScript Cloud & Infrastructure: AWS (we use a lot of it -- ECS, Lambda, S3, SQS/SNS, Bedrock, Step Functions, and more), Terraform Data: PostgreSQL, Redis, Snowflake, DynamoDB, vector databases ● AI/ML: LLMs, diffusion models, computer vision, NLP, speech-to-text -- deployed across AI infrastructure providers (RunPod, Replicate, Fal, and others) ● Orchestration: Temporal, Apache Airflow (Astronomer), LangGraph, Celery What You'll Do Own one or two systems end-to-end: understand the users and stakeholders who depend on them, then drive the architecture, implementation, operational health, and evolution. Drive cross-system technical decisions -- data contracts, API boundaries, shared infrastructure, and integration patterns across the ecosystem. Design and build production AI systems: retrieval pipelines, agentic workflows, generative pipelines, and classification engines with the right cost/latency/quality trade-offs. Set engineering standards through code reviews, design documents, and architectural RFCs. Your work becomes the reference implementation. Partner with product, design, and stakeholders to frame problems worth solving -- then solve them end-to-end. Identify and eliminate systemic bottlenecks -- whether in architecture, tooling, process, or team velocity. Elevate the engineers around you through pairing, review rigor, and raising the bar on what "production-ready" means. You Should Have Deep experience building, operating, and evolving production software systems (typically 8+ years) Track record of owning systems -- not just features -- and making architectural decisions that held up over time. Strong AI/ML engineering skills: you have built production systems that use LLMs, embeddings, classification pipelines, or generative models -- not just prototypes. Excellent systems design instincts: distributed systems, data modeling, API design, event-driven architecture, and the judgment to know when simplicity beats sophistication. Backend depth (Python, SQL, cloud infrastructure) with enough full-stack range to move confidently anywhere in the codebase. Even Better If You Bring Product sense and user empathy -- you care about whether the system actually solves the problem, not just whether it runs. Communication that scales: you write clear design docs, give precise code reviews, and can explain complex trade-offs to both engineers and non-engineers. Comfort with ambiguity and R&D pace. Some of what we build has no precedent -- you find that energizing, not unsettling.