Nidhi Lamba
CIBC
Toronto,ON
Summary
Senior AI Scientist leading end-to-end AI and machine learning delivery from business discovery through production deployment and SLA-based support. Translates BRDs, SME input, and process walkthroughs into practical solutions, then strengthens adoption through Databricks, Azure, and API-based productization. Improves production stability with model validation, monitoring, and cross-functional troubleshooting.
7
Years of experience
Work History
Senior AI Scientist
2 Years 9 Months- Supported projects through all stages of the lifecycle including business discovery, development, implementation, production deployment, post-production monitoring, and SLA-based support.
- Developed proof-of-concepts, presented findings to business and technical stakeholders, and upon approval, led implementation of solutions into production environments — demonstrating strong end-to-end delivery ownership.
- Mentored and trained team members on project context, development approaches, implementation best practices, and production readiness, contributing to stronger team capability and delivery quality across initiatives.
- Monitored deployed AI solutions in production, troubleshot failures and unexpected behaviors within SLA timelines, and coordinated with cross-functional teams to restore stability and maintain business continuity.
- Led evaluation of Databricks as a strategic platform for future AI/ML development by replicating existing VM-based solution workflows and assessing improvements in scalability, maintainability, orchestration, and model performance — helping establish a stronger foundation for enterprise-ready AI development.
- Led Databricks-based application development to improve end-user access to an internal AI solution, bridging the gap between technical development and practical business usability and supporting broader stakeholder testing and adoption.
- Supported the transformation of a large AI initiative from a project into a reusable enterprise product by identifying repeatable tasks, modularizing core functionality, and exposing capabilities through APIs — enabling multiple lines of business to evaluate and adopt the solution independently.
- Developed SDK-style wrapper functions around extrenal APIs to simplify service consumption, reduce integration complexity, and provide a consistent and reusable developer experience for the team
- Completed model validation using multiple evaluation functions and performance metrics to systematically assess output quality, reliability, and production readiness — providing structured evidence of model performance to support governance and stakeholder confidence.
- Participated in data governance and business alignment discussions, defined data requirements, supported project approvals, and quantified operational impact of AI on key business processes.
- Developed an intelligent document processing solution using Azure Form Recognizer and OpenAI prompt engineering to extract data from highly variable form layouts, reducing manual effort by over 70%. Built a Streamlit UI for real-time demos and implemented a human-in-the-loop feedback system that evolved into a low-touch workflow. Integrated the solution with internal microservices using FastAPI and conducted Swagger-based API reviews.
- Built a rule-based and ML-supported AML transaction monitoring system for FINTRAC compliance, automating the flagging of high-risk EFTs using address and name verification logic.
Data Scientist
4 Years- Led reconciliation of 15M+ entities between Panjiva and Capital IQ using BERT embeddings and fuzzy matching logic, achieving significant improvement in entity resolution accuracy and reducing false positives at scale.
- Designed an address standardization and fuzzy matching pipeline using CountVectorizer and cosine similarity, enabling accurate cross-system reconciliation across lines of business and preventing $85K in annual missed premiums.
- Developed a POC using PyTesseract, SpaCy, and NLTK to extract and identify PII in scanned documents, enhancing model accuracy through image pre-processing and post-extraction rule logic.
- Automated complex Excel and Tableau reporting workflows using Python-based data transformation scripts and fuzzy matching, reducing over 4,000 hours of manual effort annually.
- Developed Power BI dashboards analyzing sick leave patterns and work-from-home trends, enabling HR leadership to make data-driven staffing decisions.
- Ensured data consistency and quality during the SNL to CIQ Pro platform migration by developing automated QA scripts to validate transformation rules and flag discrepancies.
- Acted as liaison between analytics, product, and IT teams to document and translate business requirements into actionable data workflows, supporting cross-functional delivery.
Education
Master's - Economics
- Conducted primary data collection studying the socio-economic impacts of demonetization.
- Completed an internship with the Reserve Bank of India (RBI) analyzing credit linkages among RSETI trainees.
Bachelor's - Economics
- Relevant coursework: Game Theory, Econometrics, Behavioral Economics, Welfare Economics.
- Served as Class Representative and Joint Secretary of the Economics Society.
Skills
SQL
Databricks
MLflow
FastAPI
Power BI
Feature engineering
Model validation & evaluation
Model monitoring & drift detection
Hyperparameter tuning
Prompt engineering
BRD/SRD analysis
POC development
Stakeholder management
Data governance
Production support & SLA management
Certification
- AI-900: Microsoft Azure AI Fundamentals
- Tableau Data Scientist
- Machine Learning — Andrew NG (Coursera)
- Time Series Analysis Using Python
- SQL Essential Training
Timeline
Senior AI Scientist
CIBC
11.2023 - Current
Data Scientist
S&P Global
03.2019 - 03.2023
Bachelor's from Economics
Master's from Economics