Causal Inference for Machine Learning Engineers: A Practical Guide
- For: ML engineers and data scientists in financial services
- You'll get: Practical DAGs, propensity score recipes, and examples that plug into deep learning workflows
Durai
Rajamanickam
Shipped 100+ AI agents and governed 30+ models under ERISA, SOX, HIPAA, and fair lending.
I help boards and C-suites turn AI ambition into enterprise value — with governance that passes regulators and systems that actually ship.
Who I work with
- Banks, SMEs & regional institutions
- High-growth fintechs
- Healthcare payors
- PE-backed platforms
Currently exploring
Causal inference in financial services — how counterfactual reasoning reshapes credit risk, fairness, and regulatory decisions.
Strategic Advisor
Researcher
Author
Point of View
What I stand for
01
AI governance is not a compliance checkbox — it's your fastest path to board-level confidence.
From brake to accelerator: I build responsible AI frameworks that let leadership move faster because the risk envelope is clear, defensible, and already regulator-tested.
Banks, insurers, PE-backed platforms
02
If your CFO can't measure AI ROI, your AI strategy doesn't exist yet.
Tie every AI dollar to the P&L — with investment theses, unit economics, and post-deployment measurement that finance teams can stand behind.
CFOs, finance committees, boards
03
The hardest part of enterprise AI isn't the technology. It's building the organization that can absorb it.
I've scaled AI teams from 15 to 300+. The bottleneck is never compute or models — it's literacy, incentives, and operating structure. I build the scaffolding that makes adoption inevitable.
04
In regulated industries, the gap between a demo and production is where most AI programs die.
I've seen where AI programs typically stall between pilot and production in financial services — and I know how to close that gap under ERISA, SOX, HIPAA, and fair lending.
Books
Published Works
The AI Inflection Point: How AI Is Transforming Financial Services
- For: C-suite leaders, AI program managers in financial services
- You'll get: 7 real-world case studies (Ramp, Nubank, RBC, Stripe), a 90-day pilot plan, and ROI worksheets grounded in real numbers
Research
Research & Preprints
Peer-reviewed and preprint work at the intersection of causal ML, fairness, and large-scale data systems in financial services.
01
Decomposing Discrimination: Causal Mediation Analysis for AI-Driven Credit Decisions
Develops a doubly-robust estimator applied to ~90,000 mortgage applications, revealing that 77% of the racial denial gap operates through financial mediators shaped by structural inequality.
Read on arXiv02
Improving Legal Entity Recognition Using a Hybrid Transformer Model and Semantic Filtering Approach
Hybrid model enhancing Legal-BERT through semantic similarity filtering. Achieves 93.4% F1 on 15,000 annotated legal documents.
Read on arXiv03
Enhancing Real-Time Master Data Management with Complex Match and Merge Algorithms
Novel MDM algorithm achieving 90% accuracy on 10M+ records with 30% latency improvement, using PySpark and Databricks with Delta Lake.
Read on arXiv
Writing
Latest Thinking
Join leaders getting weekly practical briefs on enterprise AI strategy, ROI frameworks, and what's actually working in regulated industries.
The Agentic Edge — Substack
Mar 2026
The Strategic Imperative: Redefining IT Services Delivery in the Age of AI
Why IT service firms that ignore agentic AI won't survive the next 3 years.
Read →Mar 2026
How RBC's NOMI AI Delivered 17x ROI in a Regulated Industry
How compliance-first design turned regulation into competitive advantage.
Read →Start here Mar 2026
How the Top 20% of Enterprise AI Teams Deliver ROI
Patterns that separate transformational success from the rest.
Read →Mar 2026
Why Your CFO Can't Measure AI ROI (And What To Do About It)
Only 11% of CFOs can measure AI ROI — here's how to fix that.
Read →Medium
Nov 2025
Unleash the Upside: Why AI Governance is the Key to Your Firm's Next $2 Trillion Advantage
On deploying autonomous agents securely as the fastest path to outsized ROI.
Read →Aug 2025
CausalLLM: Beyond Correlation to True Causation in AI-Driven Decision Making
How open-source causal inference is transforming financial risk assessment.
Read →
Advisory
How I Work With Organizations
I take on a limited number of advisory relationships per year — typically with financial institutions, fintechs, and healthcare organizations navigating enterprise AI at scale.
Board & C-Suite Advisory
Stress-test AI strategy, shape investment theses, and build governance structures that give boards confidence to move.
- Clarify a 2–3 year AI investment thesis
- Establish a regulatory-safe governance model
- Design a 90-day path from pilot to production
Retainer · Project-based · Board sessions
Executive Briefings & Speaking
Private briefings and industry panels on AI governance, agentic systems, and the economics of enterprise AI in regulated industries.
- Executive offsites and leadership retreats
- Board education on AI risk and opportunity
Private briefings · Industry panels · Offsites
Research & Executive Training
Collaborative research on causal ML and algorithmic fairness. AI literacy programs that bring leadership teams into a shared operating language.
- Causal ML and fairness research partnerships
- Team-wide AI literacy programs
Training programs · Research partnerships
Best fit if…
- You operate in a regulated industry and need AI governance that satisfies examiners
- You have pilots but struggle to scale AI into production
- Your board is asking hard questions about AI risk and ROI
Connect
Start a Conversation
For advisory, speaking, or research collaborations, send 3–4 lines on your organization and what you're exploring.