Lead AI/ML Engineer - NextMe (Enterprise AI Insights Platform and Related AI Solutions)
Duration: June 2022 - March 2026
Summary:- Architected an enterprise AI insights platform integrating transactional, support, and behavioral data into reusable machine learning pipelines for multiple business teams
- Developed various AI solutions including knowledge assistants, document classification, recommendation services, forecasting pipelines, and computer vision workflows to enhance business operations and decision-making
Responsibilities:- Designed and implemented machine learning pipelines and AI services across multiple domains including NLP, recommendation, forecasting, and computer vision.
- Developed retrieval-augmented knowledge assistant using transformer embeddings and LLM orchestration.
- Fine-tuned transformer models for document classification, sentiment analysis, and intent detection.
- Built recommendation and forecasting services to improve user engagement and operational planning.
- Implemented computer vision workflows for product image quality and defect recognition.
- Established MLOps standards including MLflow, model registry, experiment tracking, and production monitoring.
- Deployed scalable inference APIs using FastAPI, Docker, and Kubernetes with optimized latency and autoscaling.
- Introduced drift detection, explainability techniques, and post-deployment score analysis to maintain model trust and stability.
- Collaborated with product managers, data engineers, and stakeholders to define success metrics and delivery priorities.
- Mentored engineers on feature design, error analysis, model debugging, and reproducible experimentation.
Technologies: Python, PyTorch, TensorFlow, Hugging Face, FastAPI, Docker, Kubernetes, MLflow, BERT, RoBERTa, CNNs, transformer embeddings, vector indexing, semantic search, LLM orchestration
Senior Machine Learning Engineer - Arity (Predictive Maintenance, Fraud Detection, Dynamic Pricing, and NLP Workflows)
Duration: May 2020 - May 2022
Summary:- Led development of predictive maintenance models on industrial sensor data, real-time fraud detection frameworks, dynamic pricing engines, and NLP workflows for ticket routing and issue summarization
- Designed feature stores and streaming inference pipelines to support near real-time alerting and case management
Responsibilities:- Led predictive maintenance modeling using LSTM networks and anomaly detection on sensor streams.
- Built real-time fraud detection framework with ensemble learning and risk score calibration.
- Developed dynamic pricing engine using reinforcement learning and contextual decision policies.
- Created NLP workflows for automated ticket routing, categorization, and summarization using transformer models.
- Designed feature stores and reusable preprocessing layers for consistent offline and online model behavior.
- Implemented Kafka and Spark streaming inference pipelines for near real-time scoring.
- Conducted model experimentation with cross-validation, Bayesian optimization, and error slicing.
- Built active learning loops with annotation teams to improve supervised model iteration speed.
- Containerized model services and integrated CI/CD workflows for automated deployment.
- Developed business-facing dashboards to communicate model health and outcomes.
- Provided technical leadership during production incidents and shipped corrective fixes.
Technologies: Python, LSTM, ensemble learning, reinforcement learning, transformer-based NLP models, Kafka, Spark, CI/CD, containerization
Machine Learning Engineer - ModelOp (NLP Pipelines, Customer Churn Models, Deep Learning Sentiment Analysis, OCR, and Recommendation Systems)
Duration: January 2018 - April 2020
Summary: Engineered NLP pipelines for various text data, developed customer churn and propensity models, implemented deep learning solutions for sentiment and intent recognition, created OCR and document understanding workflows, and built recommendation algorithms for content and product ranking.
Responsibilities:- Developed NLP pipelines using embeddings, topic modeling, NER, and text classification.
- Built customer churn and propensity models with gradient boosting and logistic regression.
- Implemented deep learning models for sentiment analysis and intent recognition.
- Created OCR and document understanding workflows combining image preprocessing and ML-based field interpretation.
- Developed recommendation algorithms using collaborative filtering and similarity search.
- Automated training pipelines with Airflow and Python for scheduled execution.
- Built Flask APIs to expose model predictions to internal products.
- Conducted exploratory data analysis and user segmentation using clustering and dimensionality reduction.
- Collaborated on experiment design and model evaluation against baselines.
- Documented model assumptions and deployment dependencies for maintainability.
Technologies: Python, embeddings, topic modeling, NER, gradient boosting, logistic regression, recurrent networks, transformers, OCR, Airflow, Flask
Junior AI Engineer - 3Cloud (Computer Vision Prototypes and Baseline Models)
Duration: July 2016 - December 2017
Summary: Supported early computer vision prototype development for automated quality inspection, implemented baseline machine learning models, and contributed to anomaly detection workflows and integration of inference outputs into applications.
Responsibilities:- Prepared labeled image datasets and tuned OpenCV preprocessing for computer vision prototypes.
- Built Python scripts for data cleaning and feature extraction.
- Implemented baseline models including logistic regression, decision trees, random forests, and SVMs.
- Assisted with TensorFlow and Keras training workflows and model comparisons.
- Contributed to anomaly detection workflows analyzing sensor and log data.
- Collaborated with software engineers to integrate inference outputs into web applications.
- Created internal dashboards for model evaluation metrics and data quality.
- Helped package prototype solutions into reproducible environments.
- Participated in model debugging and post-test reviews.
- Contributed to documentation and client-facing technical explanations.
Technologies: Python, OpenCV, TensorFlow, Keras, logistic regression, decision trees, random forests, SVMs, virtual environments, Git
Upstaffer since April, 2026
