AI Systems Engineer at Kulfi Collective

AI Systems Engineer

Role   AI Systems Engineer
Type   Full-Time
Location   Worli, Mumbai  (Hybrid)

THE KIND OF WORK YOU'LL SHAPE

Kulfi Collective operates at the intersection of creativity and technology, working with global brands like Spotify, Meta AI, Duolingo, and more. We’re not experimenting with AI – we’re already running agentic systems used daily by teams and clients.
Now we’re looking for a first-principles engineer to help scale what we’ve built:

• Backend systems powering AI agents
• Workflow orchestration
• Data connectors and database architecture
• Infrastructure that holds it all together

THE ROLE

As part of the tech team, you’ll build AI-powered products and workflows focussed on transforming creative and marketing processes. You’ll develop innovative digital products and AI-driven market intelligence and automation platforms.
You’ll write code alongside AI coding assistants as a primary development tool and build systems designed for other AI agents to consume.

WHAT YOU'LL DRIVE

• Build agent workflows that handle failure, retries, and real-world messiness
• Design data connectors that let agents safely discover and query brand intelligence
• Write clean, self-explanatory schemas and data migrations (readable by humans and machines)
• Instrument everything: LLM calls and agent runs, structured traces, eval harnesses, cost and latency budgets
• Maintain and improve CI/CD pipelines, containerised deployments, and observability across managed and self-hosted infrastructure
• Work within enforced-quality frameworks: automated development and infra workflows. The machines enforce the standards. You build things
• Use AI coding assistants (e.g. Cursor, Claude Code, Codex) daily as primary development tools, not supplements
• Mentor junior engineers and help establish coding and documentation best practices
• Raise the bar on code + thinking

WHAT MAKES THIS ROLE DIFFERENT

• Direct learnings from a lean engineering team that has designed and built agentic systems
• Access to a mature governance framework and methodology that lets a small team ship with disproportionate output
• Hands-on experience building agent-first infrastructure where AI is the operating model, not a feature
• Work on projects that blend creativity, culture, and code, impacting the biggest brands in India and globally
• The domain is learnable. We’ll teach it. You bring the craft

WHO YOU ARE

• 2–3 years of experience building backend systems that have actually gone live
• Strong Python/TypeScript skills (async, type hints, modern tooling)
• You write code that is clear, structured, and interpretable, by humans and AI
• Database fluency: schema design, migrations, query optimisation
• Infrastructure comfort: Docker, containers, CI/CD, debugging production systems
• Sharp eye for detail: you catch inconsistent column names, misleading error messages, and subtle bugs during review
• AI assistant fluency: hands-on experience building with tools like Claude Code, Cursor, etc., with clear judgment on when to trust outputs vs rely on core engineering thinking
• Bonus if you’ve worked on: agent systems and workflow orchestration, building agents (including browser agents), data architectures for programmatic use, observability (structured logs, tracing, metrics), and self-hosted or custom infra setups

BEFORE YOU APPLY

This role is built for someone who has put backend systems into production and watched them hold up. It’s a great fit if:

• You can point to systems you’ve built that are live and being used right now
• You already build with AI coding assistants daily, and have clear views on when to trust them and when not to
• You sweat the details others skim past: schema names, error messages, the subtle bug in review

HOW TO APPLY

Send us an email with links to things you’ve actually built. Repos, systems, agent workflows, infra you’re proud of. We care less about brand names, more about how it was made and why.