The Senior IBM AI/ML Engineer is the platform intelligence specialist on the Experis delivery team. This role owns the watsonx layer data profiling, governance scoring logic, privacy regulation classification, and LLM prompt engineering and is responsible for the quality of the governance engine's AI-powered outputs.
The primary focus for this role is taking IBM's Phase 1 pilot governance service and extending it into a production-ready Python microservice with expanded regulation coverage (HIPAA, GDPR, CCPA, FERPA, ITAR, COPPA), improved scoring logic, and clean structured JSON output consumable by Data Vault's frontend. In the back half of the engagement, this role is expected to contribute to the dynamic valuation intelligence layer, incorporating external market signals into pricing accuracy improvements.
This is a hands-on engineering role. The right candidate does not advise on watsonx they build with it. Production experience with IBM watsonx.data intelligence and IBM watsonx.ai is a hard requirement, not a nice-to-have.
Key Responsibilities
- Own the IBM watsonx.data intelligence (IBM Knowledge Catalog) integration data profiling, column-level detection, IKC asset management, and enforcement of asset cleanup mechanisms across all new components
- Design and build the regulation mapping engine extending coverage to HIPAA, GDPR, CCPA, FERPA, ITAR, and COPPA with accurate flagging logic and structured JSON output
- Develop and iterate governance scoring logic producing a consistent, defensible governance score from input data attributes, onboarding questionnaire responses, and privacy policy content
- Engineer and optimize LLM prompts against Llama, Mistral, or comparable models hosted on watsonx.ai including prompt versioning, output validation, and regression testing
- Collaborate with IBM technical leads on watsonx.ai model configuration and prompt parameter decisions
- Contribute to IBM watsonx Orchestrate workflow design for agent-consumable output formats in the later phase of the engagement
- Build and maintain governance engine unit and integration test suites using sample datasets provided by the client
- Contribute to the initial codebase assessment specifically evaluate the AI/ML layer of IBM's existing Phase 1 service for production-readiness
- Document all prompt engineering decisions, regulation classification logic, and scoring methodology to support Data Vault's team in maintaining and extending the platform after the engagement concludes
Required Qualifications
- 5+ years of hands-on AI/ML engineering experience with a strong Python background
- Production experience with IBM watsonx.data intelligence (IBM Knowledge Catalog) data profiling, column detection, and IKC asset lifecycle management; familiarity alone is not sufficient
- Production experience with IBM watsonx.ai building, testing, and deploying LLM-backed services in a non-demo environment
- Demonstrated prompt engineering experience against Llama, Mistral, or equivalent open-weight LLMs, including structured output validation and regression test strategies
- Experience building rule-based or ML-augmented classification systems regulation classification, data categorization, or comparable compliance-adjacent tooling strongly preferred
- Production-quality Python engineering skills not notebook prototyping
- Experience working within IBM Cloud environments watsonx platform specifically; general cloud AI experience on AWS or Google Cloud Platform is not a substitute
- Git version control with collaborative branching and pull request workflow experience
Preferred Qualifications
- Domain experience in data privacy, data governance, or regulatory compliance tooling familiarity with HIPAA, GDPR, CCPA, FERPA, ITAR, or COPPA classification a strong plus
- Experience with IBM watsonx Orchestrate agent workflow design and task orchestration
- Experience building AI-friendly export formats for downstream agent consumption
- Background in data valuation, data marketplace pricing, or commercial data products
Technical Skills & Platform Coverage
IBM watsonx.data intelligence (IKC) | | IBM watsonx.ai (Prompt Engineering) | | IBM watsonx Orchestrate |
|
LLM Prompting Llama / Mistral | | Regulation Classification Logic | | Python (Production-Grade) |
|
Governance Scoring Design | | IKC Asset Lifecycle Management | | Structured JSON Output Design |
|
IBM Cloud AI Platform | | Unit & Integration Testing | | Git Version Control |
What Success Looks Like
- Governance engine produces accurate regulation mappings across HIPAA, GDPR, CCPA, FERPA, ITAR, and COPPA with a structured, frontend-consumable JSON output by the engagement midpoint
- LLM prompt engineering produces consistent, validated governance outputs no hallucinated regulation flags in demo-ready deliverables
- IKC data profiling is correctly integrated and asset cleanup mechanisms are preserved and enforced across all new components
- Governance scoring logic is documented in sufficient detail that Data Vault's team can maintain and extend it after the engagement ends
- Initial codebase assessment delivers a clear evaluation of the AI/ML layer what IBM built, what works, what needs rework, and the recommended path forward
Working Environment
All environments development, staging, and production are hosted within Data Vault's IBM Cloud account. Experis engineers build and deploy within those environments; they do not own or operate cloud infrastructure independently.
The team works in a structured cadence with twice-weekly technical working sessions alongside Data Vault stakeholders and IBM technical leads. Progress is reviewed regularly through working demos and written executive updates.
This is a build-and-iterate engagement. The team is expected to move at delivery velocity while remaining responsive to product feedback as Data Vault begins demonstrating the platform to its end clients.
Never miss an opportunity! Create an alert based on the job you applied for.
