Job Role: AI & Data Intelligence Engineer
Job Type: Contract
Duration: 12 months
Work mode: Hybrid
Location: Mumbai
About the Opportunity
One of our global clients is building an enterprise platform designed to bring greater rigor, transparency, and consistency to job architecture, compensation, and pay equity.
The platform combines frontier AI, statistical modeling, and domain expertise to help organizations make better, more consistent, and defensible workforce decisions. It is being designed with a strong emphasis on accuracy, explainability, auditability, and responsible AI.
This is an opportunity to work at the intersection of generative AI, statistical engineering, and enterprise technology, building AI workflows where reliability and precision are critical to the product.
The Role
The client is looking for a specialized AI & Data Intelligence Engineer to lead the accuracy, prompt architecture, evaluation, and statistical methodology engine powering the platform.
You will work closely with the data architect, development team, and domain experts to build and maintain the AI and statistical systems behind the platform. Your primary focus will be ensuring that the platform's AI workflows and statistical models are reliable, deterministic, resistant to drift, and capable of meeting rigorous enterprise and regulatory standards.
Key Responsibilities
LLM Engineering & AI Guardrails
- Author, maintain, enhance, and version-control prompt templates.
- Design and implement robust guardrails for structured LLM outputs using JSON Schema and Pydantic.
- Build mechanisms to minimize hallucinations and ensure reliable, traceable outputs.
- Design automated retry and human-in-the-loop workflows for low-confidence or failed outputs.
- Manage the AI Gateway to capture token usage, model parameters, confidence scores, and evidence citations.
- Work with frontier LLMs, including Anthropic Claude, deployed through Microsoft Foundry.
Statistical Analytics & Verification
- Co-own and maintain the pure-Python statistical analytics engine with the data architect.
- Implement and maintain statistical methodologies supporting the platform's core analytics.
- Work with Python, pandas, statsmodels, scipy, and NumPy for OLS regression, log transformations, Gelbach decomposition, robust covariance estimation, and related statistical analysis.
- Build statistical validation and testing frameworks to compare production outputs against reference datasets and industry/market data.
- Develop backtesting and automated statistical-difference testing to ensure consistency and accuracy.
AI Evaluation, Reliability & Governance
- Build automated evaluation harnesses to measure model performance, drift, citation accuracy, and hallucination resistance.
- Use tools such as pytest, DeepEval, Promptfoo, and custom evaluation frameworks.
- Maintain version-controlled prompt templates, methodologies, and model parameters.
- Ensure a clear architectural boundary where AI handles natural-language parsing and drafting, while deterministic code performs arithmetic, scoring, and legal threshold calculations.
- Support auditability and responsible AI requirements, including relevant EU AI Act and pay-transparency considerations.
Technology StackAI & Inference
- Anthropic Claude
- Microsoft Foundry / Azure
Prompting & Structured Outputs
- JSON Schema
- Pydantic
- Jinja2
Statistical & Scientific Python
- Python 3.11+
- pandas
- NumPy
- SciPy
- statsmodels
Testing & Evaluation
- pytest
- DeepEval
- Promptfoo
- Automated statistical evaluation and diff harnesses
Version Control
Shared Application Touchpoints
- Celery
- Redis
- PostgreSQL
- Python ORM
Out of Scope / Owned by the Development Team
- Django / Django REST Framework
- Vue 3 / Vite
- nginx and web infrastructure
- Azure networking and storage provisioning
- Docker orchestration
- GitHub Actions deployment pipelines
Must have
- Bachelor’s or Master’s degree in Data Science, Computer Science, AI/ML, Quantitative Economics, or a related quantitative/technical discipline.
- Hands-on experience building and orchestrating frontier LLM applications.
- Strong experience with structured LLM outputs, JSON Schema, and Pydantic.
- Strong Python programming and statistical modeling experience.
- Hands-on experience with pandas, statsmodels, scipy, and NumPy.
- Experience building automated testing and evaluation frameworks for AI/ML systems.
- Understanding of model reliability, drift, hallucination mitigation, and output validation.
- Strong ability to work across AI engineering, statistical modeling, and product requirements.
Preferred
- Experience with Anthropic Claude and/or Microsoft Foundry/Azure AI.
- Experience in compensation analytics, job evaluation, pay equity modeling, or workforce analytics.
- Familiarity with the EU AI Act, EU Pay Transparency Directive, or similar regulatory frameworks.
- Experience building enterprise-grade AI systems where accuracy, traceability, and explainability are critical.
Why This Opportunity
- Work directly with frontier AI models in a production enterprise application.
- Build AI systems where accuracy, reliability, explainability, and responsible AI are core to the product.
- Combine LLM engineering with real-world statistical modeling rather than working solely on generic AI applications.
- Work closely with data, engineering, and domain experts to shape an emerging AI-powered product.
- Have a direct impact on how organizations approach job architecture, compensation, and pay equity.
- Opportunity for full-time conversion following the initial 12-month engagement.
Pay: ₹1,200,000.00 - ₹1,800,000.00 per year
Application Question(s):
- How many years of professional experience do you have in AI/ML, Generative AI, or LLM engineering?
- How many years of hands-on Python experience do you have, particularly in statistical modeling/data analysis?
- Have you worked with structured LLM outputs using JSON Schema and/or Pydantic?
- This is a 12-month contract position. Would you be comfortable with the contract duration?
Work Location: Hybrid remote in Mumbai, Maharashtra (Mumbai, Mumbai Suburban District)
