Overview
Lead the technical direction and end-to-end ownership of an enterprise AI platform, covering a knowledge graph, secure agentic AI systems and the supporting cloud, data and delivery infrastructure, used across drug discovery, development and operations.
What you'll do
- Operate and evolve the knowledge graph and graph database platform to enterprise standards, ensuring performance, availability, backup/restore, access management and smooth operations.
- Make graph and other data stores readily usable by AI agents through secure tools, accelerating insight generation.
- Define and implement secure agent orchestration through gateways, and build and maintain MCP servers and tool layers with clear guardrails.
- Establish reusable patterns for prompt, skill and tool management, and continuously evaluate agents with automated benchmarks to improve quality, cost and latency.
- Provide a consistent, secure access layer to AI tools and internal applications, maintaining shared sign-in services, protecting credentials and enforcing authorization.
- Deploy and run workloads on Kubernetes using Helm with robust ingress, secrets management, storage sizing and per-pull-request review environments.
- Provision infrastructure as code across Azure, AWS and GCP to keep deployments portable and resilient.
- Build and operate reliable, resumable bulk ingestion pipelines for multi-terabyte datasets across graph, analytical and search stores, processing large Parquet datasets efficiently and managing deltas and staged releases.
- Design and maintain GitHub Actions pipelines for build, test, review apps and deployment (including self-hosted runners), and run deployments using GitOps.
- Build and maintain backend services in Python and frontends in React/TypeScript with strong authentication, validation, pagination and versioning, owning services end to end.
- Set platform priorities and direction, review designs and pull requests, mentor engineers, and partner with data, NLP, application, security and identity teams.
- Create architecture documents, runbooks and integration guides, and communicate designs clearly to technical and non-technical audiences.
What you'll need
- 8+ years in platform, DevOps, SRE or infrastructure engineering, including experience leading technical work or mentoring others.
- Working knowledge of knowledge graphs, and hands-on experience with graph databases such as Neo4J.
- Experience writing skills and building gateway-based agent orchestration for LLM-based assistants.
- Expertise in processing Parquet data at scale (with PyArrow, pandas, Polars or Spark), including partitioning, schema handling and memory-efficient loading.
- Full-stack development experience: Python with FastAPI for REST APIs, React with TypeScript for frontends, and the ability to maintain and upgrade existing applications and dependencies.
- Ability to collaborate effectively with Databricks platform and data teams, including integrating with data they produce.
- Strong Python and shell scripting for automation and tooling.
- Deep experience with Kubernetes and Helm, and hands-on experience across Azure, AWS and GCP.
- Good experience with CI/CD, GitHub Actions, and GitOps.
- Hands-on experience implementing single sign-on MSAL against Microsoft Entra ID, both frontend (React) and backend applications.
- Solid grounding in authentication and authorization: OAuth 2.0/OIDC, JWT, Microsoft Entra ID and API gateway concepts.
- Experience running data stores in production.
- Infrastructure as code (Bicep, Terraform or similar) and sound secrets management.
- Clear written communication and good documentation habits.
Nice to have
- Hands-on experience building Model Context Protocol (MCP) servers or AI gateways.
- Experience with API gateways and proxies.
- Large-scale data ingestion (multi-terabyte) and database performance tuning.
- Observability tooling.
- Interest in life sciences or biomedical data.
Details
- Location: Chennai, Tamil Nādu, India.
- Work mode: in-person, a minimum of three days per week from the office.
Read the full description and apply on the company’s own careers page.