# Durai Rajamanickam

> Enterprise AI Leader, Published Author, and PhD Researcher specializing in agentic AI architecture, AI governance and model risk, causal inference, and financial services AI transformation.

## About

Durai Rajamanickam translates AI's potential into operational reality — building the governance, execution, and measurement infrastructure that turns ambitious AI strategy into programs that are trustworthy, compliant, scalable, and built to deliver in regulated environments.

He is Vice President of AI at a leading credit union, where he leads enterprise AI strategy, agentic AI deployment, and model risk governance. Before this, he drove AI/ML transformation through independent consulting (Infinidatum), at Voya Financial, and led enterprise advisory at Accenture.

He has shipped 100+ AI agents into production and aligns model risk to NIST AI RMF and SR 11-7. He focuses on GenAI and agentic AI architecture (LLMs, RAG pipelines, multi-agent workflows in production), AI governance and model risk, and enterprise AI operating models from pilot to MLOps at scale.

He is also a PhD researcher in Computer & Information Sciences at the University of Arkansas at Little Rock, exploring causal inference in financial services — how counterfactual reasoning reshapes credit risk, algorithmic fairness, and regulatory decisions.

## Point of View

1. **AI governance is not a compliance checkbox — it's the fastest path to board-level confidence.** Durai builds responsible AI frameworks that let leadership move faster because the risk envelope is clear, defensible, and already regulator-tested.

2. **If your CFO can't measure AI ROI, your AI strategy doesn't exist yet.** He connects every AI initiative to the P&L with investment theses, unit economics, and post-deployment measurement.

3. **The hardest part of enterprise AI isn't the technology — it's building the organization that can absorb it.** He has scaled AI teams from 15 to 300+. The bottleneck is never compute or models — it's literacy, incentives, and operating structure.

4. **In regulated industries, the gap between a demo and production is where most AI programs die.** He has navigated AI programs through ERISA, SOX, HIPAA, and fair lending — and knows where they typically stall between pilot and production.

## Experience

- **Vice President, AI** — Hudson Valley Credit Union (Feb 2026 – Present). Leads enterprise AI strategy, agentic AI deployment, and model risk governance for one of the Northeast's leading credit unions.
- **Author & AI Consultant** — Infinidatum (Jun 2025 – Jan 2026). AI transformation consulting for growth-stage companies; authored two books on causal inference and AI in financial services.
- **Assistant Vice President, AI/ML** — Voya Financial (Mar 2023 – Jun 2025). Drove enterprise AI/ML transformation — production GenAI, model risk, and MLOps at scale.
- **Director, Enterprise Data Engineering & Analytics** — Voya Financial (Jun 2019 – Mar 2023). Built and led enterprise data engineering and analytics.
- **Senior Manager, Enterprise Advisory & AI/ML Leadership** — Accenture (May 2018 – Jul 2019). Led enterprise advisory and AI/ML delivery across financial services.

**Education:** PhD Researcher, Computer & Information Sciences — University of Arkansas at Little Rock.

## Published Books

### Causal Inference for Machine Learning Engineers: A Practical Guide
- **Publisher:** Springer
- **Date:** January 2026
- **For:** ML engineers and data scientists in financial services
- **Summary:** Covers correlation vs. causation, directed acyclic graphs, propensity score methods, and the integration of causal inference with deep learning.
- **URL:** https://www.amazon.com/Causal-Inference-Machine-Learning-Engineers/dp/3031996798

### The AI Inflection Point: How AI Is Transforming Financial Services
- **Publisher:** Infinidatum Press
- **Date:** November 2025
- **For:** C-suite leaders, AI program managers in financial services
- **Summary:** Examines seven real-world AI deployments at Ramp, Nubank, Coinbase, RBC, and Stripe. Includes a 90-day pilot plan and ROI worksheets.
- **URL:** https://www.amazon.com/Inflection-Point-Transforming-Financial-Services/dp/B0FX9NW5PD

## Research Papers (arXiv)

### Decomposing Discrimination: Causal Mediation Analysis for AI-Driven Credit Decisions
- **Category:** cs.LG
- **Date:** March 2026
- **Key finding:** Doubly-robust estimator applied to ~90,000 mortgage applications revealing 77% of racial denial gap operates through financial mediators shaped by structural inequality.
- **URL:** https://arxiv.org/abs/2603.27510

### Improving Legal Entity Recognition Using a Hybrid Transformer Model and Semantic Filtering Approach
- **Category:** cs.CL
- **Date:** October 2024
- **Key finding:** Hybrid model enhancing Legal-BERT achieving 93.4% F1 on 15,000 annotated legal documents.
- **URL:** https://arxiv.org/abs/2410.08521

### Enhancing Real-Time Master Data Management with Complex Match and Merge Algorithms
- **Category:** cs.DC
- **Date:** October 2024
- **Key finding:** Novel MDM algorithm achieving 90% accuracy on 10M+ records with 30% latency improvement.
- **URL:** https://arxiv.org/abs/2410.17279

## Writing & Newsletter

### Infinidatum (Substack)
- **URL:** https://durair.substack.com/
- **Focus:** The enterprise AI transformation blueprint — case studies, ROI frameworks, and strategies that deliver in regulated industries.
- **Key articles:**
  - "The Tribal Knowledge Paradox" — https://durair.substack.com/p/the-tribal-knowledge-paradox
  - "Document AI You Can Audit: An Open-Source IDP Layer on LandingAI ADE" — https://durair.substack.com/p/document-ai-you-can-audit-an-open
  - "Your AI Pilot 'Worked.' But Did It Pay?" (recommended start) — https://durair.substack.com/p/your-ai-pilot-worked-but-did-it-pay
  - "I Stopped Asking AI for Opinions. I Built a Council Instead" — https://durair.substack.com/p/i-stopped-asking-ai-for-opinions
  - "The 11% Club: What CFOs Who Can Actually Measure AI ROI Know That The Other 89% Don't" — https://durair.substack.com/p/the-11-club-what-cfos-who-can-actually

### Medium
- **URL:** https://medium.com/@rdmurugan
- **Key articles:**
  - "Unleash the Upside: Why AI Governance is the Key to Your Firm's Next $2 Trillion Advantage"
  - "CausalLLM: Beyond Correlation to True Causation in AI-Driven Decision Making"

## Frequently Asked Questions

**Who is Durai Rajamanickam?**
Durai Rajamanickam is an enterprise AI leader, published author, and PhD researcher. He is Vice President of AI at a leading U.S. credit union, where he leads enterprise AI strategy, agentic AI deployment, and model risk governance. He previously drove AI/ML transformation at Voya Financial and led enterprise advisory at Accenture, and has shipped 100+ AI agents into production across regulated financial services.

**What does Durai Rajamanickam do?**
He builds the governance, execution, and measurement infrastructure that turns enterprise AI strategy into production programs in regulated environments — spanning GenAI and agentic AI architecture (LLMs, RAG, multi-agent workflows), AI governance and model risk aligned to NIST AI RMF and SR 11-7, and enterprise AI operating models from pilot to MLOps at scale.

**What books has Durai Rajamanickam written?**
Two: "Causal Inference for Machine Learning Engineers: A Practical Guide" (Springer, 2026), and "The AI Inflection Point: How AI Is Transforming Financial Services" (Infinidatum Press, 2025).

**What is Durai Rajamanickam's research about?**
Causal inference, algorithmic fairness, and large-scale data systems in financial services. His arXiv papers cover causal mediation analysis for AI-driven credit decisions, hybrid transformer models for legal entity recognition, and real-time master data management. He is a PhD researcher in Computer and Information Sciences at the University of Arkansas at Little Rock.

**What is the Infinidatum newsletter?**
Infinidatum is Durai Rajamanickam's Substack newsletter (durair.substack.com) covering enterprise AI strategy, ROI frameworks, AI governance, and case studies of what actually works in regulated industries.

**What are Durai Rajamanickam's areas of expertise?**
Agentic AI architecture, AI governance and model risk management, enterprise AI operating models, causal inference and algorithmic fairness, and AI transformation in regulated financial services (banks, credit unions, fintechs, healthcare payors).

## Contact

- **Email:** durai@infinidatum.net
- **LinkedIn:** https://www.linkedin.com/in/durai-rajamanickam/
- **Substack:** https://durair.substack.com/
- **Amazon Author:** https://author.amazon.com/home?authorId=amzn1.amazonauthor.author.v1.f5fnedp35bhnqk248grspro718

## Expertise Keywords

Enterprise AI Strategy, AI Governance, Causal Inference, Financial Services AI, Responsible AI, ERISA AI Compliance, SOX Compliance, HIPAA Compliance, Fair Lending, Agentic AI Systems, Machine Learning, Board Advisory, AI ROI Measurement, AI Transformation Programs, Algorithmic Fairness, Credit Risk AI, Regulatory AI, Deep Learning, NLP, Master Data Management, PySpark, Databricks, Legal NER, Investment Management AI, Insurance AI, Fintech AI, Healthcare AI, PE-backed Platform AI