Cleveland, OH
Nandini
AI Architect · ML / MLOps · Gen AI & Agentic AI Engineer
I build across the full AI stack — ML & MLOps, Gen AI and agentic systems, AI backends, and cloud data platforms — and embed with teams as a forward-deployed engineer to ship enterprise-grade products.
Gen AI · MLOps · Agents
Turning complex AI into products people trust
AI Architect and full-stack AI engineer with 3+ years designing scalable, enterprise-grade AI systems and leading cross-functional engineering teams at BLK International, Two Sigma, and Northern Trust. I work end to end across ML, MLOps, Gen AI, and agentic systems — and as a forward-deployed engineer embedding with customers to ship. I translate complex AI capabilities into audit-ready products: reducing false positives by 40% through HITL validation, cutting cloud infrastructure costs by 30%, and slashing review times by 50% with internal diagnostic dashboards.
I've led engineering across BLK International, Two Sigma, and Northern Trust — shipping RAG platforms, LLM inference pipelines, and diagnostic dashboards that hold up to real compliance scrutiny.
Education
Anderson University
Master's in Computer Science
Aug 2023 – Aug 2025 · Greenville, SC
- Name
- Nandini
- Focus
- AI · ML · MLOps · Gen AI
- Degree
- M.S. Computer Science
- Location
- Cleveland, OH
- mttnandini@gmail.com
- Experience
- 3+ years
Where I've built & led
Jan 2026 – Present
BLK International LLC
Chief Technology Officer (CTO) & Lead AI Architect · Boca Raton, FL
- Architected and launched an AI-powered compliance platform using Amazon Textract and AWS Bedrock (Claude 3), reducing manual document review time by 85% and achieving 94% accuracy in regulatory clause detection for enterprise clients.
- Designed a scalable microservices ecosystem on AWS ECS (Fargate) and FastAPI, integrating SQS for asynchronous OCR processing and OpenSearch for RAG-based compliance queries, supporting burst loads of 500+ documents/hour with sub-200ms API latency.
- Led a cross-functional team of 6 engineers (Backend, Frontend, DevOps, QA) to deliver the MVP in 3 months, establishing CI/CD pipelines via GitHub Actions and defining system architecture that secured enterprise client renewals worth $250k+.
- Spearheaded AI strategy by implementing Human-in-the-Loop (HITL) validation workflows, ensuring data integrity and model trustworthiness for high-stakes financial compliance decisions, reducing false positives by 40%.
- Optimized cloud infrastructure costs by 30% through right-sizing Lambda functions and implementing efficient caching strategies with Redis/ElastiCache for frequent compliance rule lookups.
Oct 2024 – Dec 2025
Two Sigma
AI/ML Engineer · New York (Remote)
- Rectified feature pipelines and identified data quality deviations, reducing inconsistent experimental outcomes by 40% across quantitative research teams.
- Standardized ML preprocessing and evaluation pipelines enabling consistent classification, forecasting, and anomaly detection experiments with 99% reproducibility.
- Enhanced model benchmarking cycles by introducing cross-validation and automated hyperparameter tuning, increasing confidence in model robustness by 35% across configurations.
- Strengthened experiment traceability by version-controlling pipelines, organizing artifacts and metrics, and storing outputs in AWS S3.
- Orchestrated the creation of internal diagnostic dashboards, surfacing critical model performance metrics and feature sensitivity for researchers, slashing review times by 50%.
- Expedited the ML lifecycle by aligning modeling assumptions with quantitative researchers and ML engineers, improving data preparation, evaluation, and insight delivery processes by 25%.
Aug 2023 – Sep 2024
Northern Trust
ML Engineer · Cleveland, OH
- Built forecasting and classification models on electronics and communication system datasets, improving operational planning and monitoring accuracy.
- Articulated predictive model outcomes and principal variance drivers into executive-ready summaries, accelerating decision-making for risk and compliance departments by 25%.
- Formulated comprehensive Power BI and Tableau dashboards that synthesized historical trends, model outputs, and KPI variance indicators, improving operational health visibility by 40%.
- Strengthened governance and traceability by managing analytical artifacts and intermediate datasets in AWS S3, supporting secure access and compliance requirements.
- Standardized recurring analytics preprocessing and ML pipelines, cutting manual effort by 60% and enhancing reporting cycle consistency.
Selected work
FeaturedEnd-to-end RAG platform that ingests regulatory documents, extracts clauses via OCR, and answers compliance queries with human-in-the-loop validation.
Python · FastAPI · AWS S3 / Textract · OpenAI / Bedrock · LangChain · FAISS / Pinecone · React · Docker
FeaturedEnterprise Performance Monitoring Dashboard with AI Insight Layer
Interactive analytics dashboard aggregating KPIs and surfacing AI-generated insights across time-based operational trends.
Python · Pandas · NumPy · Streamlit · Plotly · Streamlit Cloud
Tesla Telemetry Monitoring System
Time-series pipeline that ingests vehicle telemetry, detects anomalies, and visualizes fleet health in real time.
Python · Pandas · NumPy · Streamlit · Jupyter
Multimodal Tumor Classification
Deep learning model fusing imaging modalities to classify tumors with augmentation-driven robustness.
Python · TensorFlow · Keras · NumPy
JPMorgan Chase Backend System Simulation
Event-driven transaction backend simulation built with a service-oriented architecture and asynchronous processing.
Java · Spring Boot · Apache Kafka · REST APIs · SQL · Spring Data JPA
Walmart Advanced Software Engineering Simulation
Software architecture exercise covering data structures, database modeling, and process optimization across Java and Python.
Java · Python · SQL · UML
What I work with
Generative AI & LLM Systems
95%- LLM Inference Pipelines
- Prompt Engineering
- RAG
- Embeddings
- Vector Similarity Search
- Model Monitoring
- Human-in-the-Loop (HITL)
ML & Predictive Modeling
92%- Supervised Learning
- Classification & Regression
- Time-Series Forecasting
- Feature Engineering
- Cross-Validation
- Hyperparameter Tuning
- LightGBM
- scikit-learn
- Anomaly Detection
- Model Drift Detection
- AutoML (H2O.ai)
System Architecture & Cloud
88%- System Design
- Microservices
- REST APIs
- Distributed Systems
- Event-Driven Architecture
- AWS (S3, EC2, Lambda, Textract, Bedrock)
- Docker
- Kubernetes
- CI/CD (GitHub Actions)
Deep Learning & NLP
85%- Transformers
- spaCy
- Embeddings
- Text Preprocessing
- CNNs
- GANs
- TensorFlow
- Keras
- SHAP
- LIME
Programming & Data
93%- Python
- SQL
- MySQL
- Pandas
- NumPy
- Linux
- Git
- GitHub
- Jupyter
- VS Code
Backend
82%- Java
- Spring Boot
- Apache Kafka
- FastAPI
- API Development
Leadership & Strategy
90%- Technical Leadership
- System Architecture Planning
- Cross-Functional Collaboration
- Agile Development
- Product Strategy
- Stakeholder Communication
Tools & technologies
The core stack I build and ship with.
Python
TensorFlow
Keras
PyTorch
scikit-learn
Hugging Face
LangChain
OpenAI
AWS
Docker
Kubernetes
FastAPI
Apache Kafka
Spring Boot
Java
MySQL
Pandas
NumPy
Streamlit
Plotly
Jupyter
Tableau
Git
GitHub
Linux
React
Certifications
- May 2026
JPMorgan Chase Software Engineering Virtual Experience
JPMorgan Chase
- Feb 2026
Walmart USA's Advanced Software Engineering
Walmart Global Tech
- Mar 2024
AWS Academy Graduate – Cloud Security Foundations
AWS Academy
- May 2019
Robotics & Embedded Systems Workshop
Technophilia Solutions, NIT Warangal
Interview me in seconds.
Ask anything you would in a screening call, or paste a job description to get an instant fit assessment. Answers are grounded in Nandini's real experience.
Ask Nandini
AI assistant · grounded in her resume
Try one of these to get started:
Ready to talk? Schedule a 30-min call with Nandini.
AI responses may contain mistakes. For anything important, contact Nandini directly.
Let's build something
intelligent.
Open to AI architecture, technical leadership, and CTO-level opportunities. The fastest way to reach me is by email.