Datagrok

The project

You will own the development of a capability for scientific models - a custom application we are building on the Datagrok platform for one of our pharma customers. It's a central place where scientists register, discover, run, and compare computational models. Models come from different teams, in different languages, with different inputs and outputs. Your job is to turn that into one coherent, fast, self-service experience.

This is not a ticket-driven role. You will "sit" with the customer's scientists, understand how they work, and translate that into working software.

Responsibilities:

• Develop clean, high-performance, well-documented, production-level code
• Own your backlog: prioritize, estimate, communicate status, and drive features from a conversation to production
• Talk directly to the scientists and computational modelers: run demos, gather requirements, challenge assumptions, propose better ways to solve their problems on the platform
• Contribute improvements back to the platform and public plugins where the project uncovers gaps

Requirements:

• Expert-level TypeScript: you've designed and shipped non-trivial TS applications or libraries in pre-LLM era
• Engineering range: building elegant solutions from scratch, evolving complex existing codebases, and contributing efficiently across the whole stack, UI to backend
• Past projects in scientific or life-sciences software, complex modeling, or scientific computing (e.g., simulation engines, computational tools, statistical modeling apps, numerical/HPC code, etc.)
• Fluent spoken and written English: you will be in customer-facing meetings weekly, presenting your work and negotiating scope with scientists
• Willingness to learn new scientific domains
• Demonstrated ability to self-manage: you've owned a project or major workstream end-to-end and can point to something that shipped because you drove it to completion

Nice to have:

• Scientific background (STEM / life sciences)
• Python, R, C++, or Julia (enough to read and integrate scientists' model code)
• Experience with ML development and/or model lifecycle tooling (model registries, MLflow-style tracking, workflow engines)

We offer:

• Fully remote work
• Excellent growth opportunities in a highly dynamic, unique product company
• Zero bureaucracy
• Paid trainings and certifications
• Competitive compensation

We're a 20-person team building a browser-native scientific data platform used by Novartis, J&J, Daiichi-Sankyo, and other top pharma and biotech companies. Our platform was already 10x faster than the competition and delivered more value for less — even before the AI era. Now, we're taking it to the next level.

We're competing with Spotfire, Certara, Dotmatics, and Schrödinger — and winning. How? By building faster and smarter. Our engineering team has fully embraced AI-driven development (Claude Code, agentic workflows, AI-assisted testing), and it's become our biggest force multiplier. We ship in weeks what used to take quarters.

We're looking for an engineering leader who's already living this way — someone who has hands-on experience with agentic development tools and can help lead a team that's pushing the boundaries of what a small, focused group can deliver. You'll report directly to the CEO and have real authority to shape how we build.

What you'll build and lead

This isn't a "manage meetings and write status reports" role. You'll own critical product areas and ship alongside the team. Here's what's on the roadmap:

• AI-powered agentic workflows — Autonomous scientific agents that utilize Datagrok's plugins and integrations as tools.
• Automated testing at scale — Automatic scenarios using Playwright + AI, AI-driven bug reproduction and fixing, stress tests, unit tests, system health tests, plugin tests - everything integrated into CI-CD and managed by AI
• Platform evolution — Cross-application integration and eventing mechanicm, ClickHouse as an analytical backbone, high-performance computations
• Enterprise integrations — Databricks connectivity (ADBC/Arrow for 2–10x speedups), HuggingFace model integration, and dashboard importers for Tableau and Spotfire migrations
• DevOps and infrastructure — ArgoCD deployments, Docker image management, containerized dev environments, and OpenTelemetry observability across a multi-layer stack
• AI-powered knowledge graph — mapping source code, features, tests, docs, and Jira tickets to enable impact analysis and agentic development workflows across the entire codebase
• Process improvements - code reviews, better planning/outcome system, DX improvements, onboarding for new developers

What we're looking for

Must have:

• You've already adopted autonomous agentic development in your own workflow (harnesses, tool-use architectures, AI-driven testing, etc)
• 5+ years as a strong engineer who still codes. This is a hands-on position
• Experience leading a team (formal title matters less than demonstrated leadership)
• Proficiency in TypeScript or Dart; comfort picking up new languages fast. Our core stack is Dart and TypeScript, with C++/WASM, Python, R, Julia, and Matlab in scientific plugins
• Track record of delivering complex software products — not just managing people who do
• You see a problem, you fix it. You don't write a proposal about fixing it. You don't wait for permission
• Strong communication skills and fluent English — you'll work directly with technical clients at pharma companies and present to enterprise IT teams

Strong pluses:

• Life sciences, cheminformatics, or bioinformatics background (or genuine curiosity about this space)
• Experience building agentic AI systems — not just prompting LLMs, but designing autonomous pipelines, harnesses, tool-use architectures, or AI-driven testing frameworks
• Background in high-performance computing, data visualization, or scientific computing
• Startup DNA — you've built things from zero, worn multiple hats, and thrive without a playbook

What we offer

• Real ownership — reporting directly to the CEO, you'll lead critical product areas for a platform deployed at top-10 pharma companies, not maintain someone else's legacy code
• An AI-first engineering culture — we're not "exploring AI adoption." We've already made the transition. You'll help take it further
• Direct client impact — your work will be used daily by scientists doing drug discovery at companies you've heard of
• Startup speed, enterprise clients — the autonomy and pace of a small team, with the validation of Fortune 500 deployments
• Fully remote, zero bureaucracy, flexible hours, unlimited PTO
• Competitive compensation with growth equity opportunity
• Paid training and certifications