Principal/Lead Engineer, Embedded Data and Agentic AI Engineering Position Principal or Lead Engineer, embedded with Baker Hughes teams Engagement Baker Hughes account, Codvo AI Location Pune, India (hybrid); travel to Baker Hughes sites as needed Employment Type Full time Reports To Head of Technology and Architecture, Codvo AI Experience 10+ years, with 3+ years on AI/ML systems in production
About the Engagement Codvo AI is expanding its engineering team for the Baker Hughes account. The engagement covers four workstreams agreed with Baker Hughes leadership: roadmap acceleration (including the Iguazio to Red Hat OpenShift AI migration), agentic AI in production, a unified data foundation, and forward deployment. Codvo brings a Red Hat Marketplace certification and delivery experience in industrial and energy accounts. Role Summary You will be Codvo's senior embedded engineer inside Baker Hughes teams, leading the agentic-AI-in-production and unified data foundation workstreams. The core technical problem is contextualization: Baker Hughes and its customers run multiple siloed asset hierarchies (SAP, health, integrity, physical location, IT structure) with no unified layer across them. You will lead the rapid POC for asset hierarchy contextualization and carry it toward production, working directly with Baker Hughes platform and central architecture leadership. Responsibilities • Act as technical lead for agentic AI use cases on the account, from POC scoping through production hardening. • Design the contextualization and semantic layer that unifies siloed asset hierarchies (SAP, health, integrity, location, IT structure). • Lead the asset hierarchy contextualization POC, applying a synthetic data approach (built from public OEM specs) so no customer data is needed to prove value. • Integrate with Databricks and Unity Catalog where the customer has them; design the standalone path where they do not (about 80 percent of Baker Hughes customers have Databricks, but not all have Unity Catalog fully configured). • Set engineering standards for the embedded team: code review, testing, release discipline, and documentation. • Work directly with Baker Hughes platform and central architecture stakeholders on architecture decisions and whiteboarding sessions. • Mentor the account's data scientists, AI engineers, and data engineers. • Feed delivery lessons back into Codvo's reusable components and connector library. Must-Have Skills and Experience • 10+ years of software or data engineering, with 3+ years designing and running AI/ML systems in production. • Multi-agent or LLM orchestration in production (LangGraph or equivalent), not only prototypes. • Deep data platform experience: Spark, Databricks, Delta Lake, and semantic or metrics layer design. • Kubernetes-based deployment experience; comfort with OpenShift specifics. • Track record as an embedded or forward-deployed engineer at enterprise customers: you can hold your own in a customer architecture review. • Ability to scope a POC in days, ship it in weeks, and state honestly what it proves and what it does not. Nice to Have • Oil and gas, energy, or heavy-industry domain experience. • Unity Catalog governance design. • Red Hat OpenShift AI or other Kubeflow-based ML platforms. • GIS or asset hierarchy data (ArcGIS, asset registries, functional location trees). • Prior work with industrial contextualization platforms (Cognite or similar). First 90 Days • Weeks 1 to 4: map Baker Hughes's hierarchy silos with their data team and agree POC scope with the account stakeholders. • Weeks 5 to 10: deliver the asset hierarchy contextualization POC on synthetic data and demo it to Baker Hughes. • Weeks 11 to 13: agree the production path (integration model, data sources, governance) and staff the follow-on plan.