Nov 2021 – Dec 2022 · 1 yr 1 mo · Singapore

• I co-instructed a part-time Data Science immersive BootCamp. I mainly cover topics related to NLP, Deep Learning and Big Data processing.
• I created new lessons on MLOps on the cloud using MLFlow, Flask, Docker and Google Cloud Run for severless cloud deployment of ML models and Streamlit on Streamlit Cloud as a UI to gather inputs from users, post requests to the API and display the model predictions returned by the API

Jun 2021 – Sep 2022 · 1 yr 3 mos · Singapore

• I manage a team of Data Scientists to research, architect, develop and scale cutting-edge Machine Learning products from concept to production using cloud technologies. Under my leadership at Edelman, the team has grown by 3X and covers the entire end-end product development lifecycle.
• Personally, I'm a hands-on leader who loves to code! I have significant expertise in the full stack of Data Science product development such as:
• 1. ETL pipelines for near real-time as well as batch big data processing
• 2. Deep Learning research, model development, tuning, and deployment for both real-time and batch processing in AWS and GCP
• 3. Serverless microservices-based backend development using Docker containers and Cloud technologies such as Lambda, API Gateway, Cloud Run, etc
• 4. Frontend development using Plotly Dash, Bootstrap, CSS and Streamlit
• 5. MLOps using AWS Sagemaker, GCP VertexAI and other technologies
• I'm also actively engaged in presenting ML products and services to clients. I love explaining complex ML concepts in simple and easy-to-understand language with lots of visuals!

Nov 2019 – Jun 2021 · 1 yr 7 mos · Singapore

• Responsible for architecting and scaling various ML/DL solutions using Google Cloud Platform tools such as AI Platform notebooks, training jobs, model endpoints, Compute Engine, Big Query, GCS, Dataflow & Dataproc.
• Developed an ultra-scalable product matching and anomaly detection service using Sentence Transformers NLP and distributed ANN technologies including Elasticsearch, Milvus, and GCP SCANN.
• Architected and deployed RESTful microservices using Docker, GCP Compute Engine, and Seldon service on GKE.
• Working-level knowledge in various Agile software development tools (Jira, Confluence, Github), CI/CD (TravisCI, Jenkins) and DevOps (Terraform, Ansible).
• Approximate Nearest Neighbors (ANN):
• Achieved >85% Recall at 98% Blocking within 500ms for 50Million product pairs and <20ms latency for K=100 NN query across 30Million unique products.
• Architected and developing an enterprise-level “Embedding Feature Store” with 10ms vector retrieval and 20ms ANN search across multiple images, text, and clickstream domains.
• Contributed Opendistro benchmarking to ANN-Benchmarks open-source GitHub repository.
• Wrote multiple medium posts on ANN topics with total ~30K views and >30stars on Github.
• Natural Language Processing (NLP):
• Experienced in various SoTA NLP techniques and frameworks including FastText, and BERT.
• Trained large-scale deep learning neural networks in distributed Tensorflow for name-gender prediction on 40Mil names with 1.5% CTR improvement and product reviews analytics on 200Mil reviews and QnA with 0.32% CVR improvement.
• Co-developed an in-house Python package for Natural Language Pre-Processing of Bahasa Indonesian and English text which is widely adopted within the internal DS community.
• Co-authored a paper that was accepted to the International ACL2020 conference in Seattle.
• Top10 code contributor to the DS GitHub repository.

Sep 2017 – Oct 2019 · 2 yrs 1 mo

• Managed team of data engineers/scientists to develop production full-stack engineering solutions including ETL, analytics, and visualization in Hadoop ecosystem, SQL, Teradata, Tableau, and Web.
• Developed showcase projects that enabled Micron to be one of Singapore’s first factories to be awarded the prestigious World Economic Forum Lighthouse Network for leadership in I4.0 technology.
• Engaged cross-functional business stakeholders to scope and propose data-driven projects for both operational efficiency improvement as well as advanced A.I. analytics.
• Own, prioritize, plan, and execute a portfolio of data science, data engineering, BI, and big data projects.
• Won the prestigious “Idea of the Quarter” award for an innovative algorithm developed that reduces 90% of Type1 errors in A/B testing data analysis and results in 80% faster decision making.
• Led several Tableau dashboarding projects to monitor and drill down into root cause of issues saving > $100M in 2018.
• Led the investigation and optimization of an HDFS, HBase, Hive, Spark, and Python data pipeline that successfully reduced missing data rate from 90% to 2% within four weeks. Implemented additional logging and built telemetry dashboards to monitor data quality continuously.
• Led a Deep Learning based image similarity to mimic human behavior while recognizing patterns that could accurately catch process root cause defects which are responsible for 30-40% yield loss issues using Python, bash, HBase, and Spark.
• Leading a Machine Learning based virtual metrology effort to reduce costly actual product measurements using Python, bash, NIFI, and HBase with est $1M+ savings.
• Developed a DOE data analysis reporting portal using Python, SQL, bash, and HTML. Coded up an analysis tool to use Natural Language Processing on job application resumes and rank top candidates based on suitability to Job Description.

Mar 2013 – Aug 2017 · 4 yrs 5 mos

• Implemented innovative test programs for quality, yield, and test time improvement.
• Developed novel data analytics automation for test results using Bash, Python, JMP Script, and HTML that reduced the time spent on data analytics from several hours to several minutes.
• Lead several analytics product integration projects using a combination of statistical data analytics including ANOVA, KW, regression, classification and on product DOE testing. Developed and monitored several KPI dashboards and automated emails.

Feb 2012 – Feb 2013 · 1 yr

Jun 2010 – Jan 2012 · 1 yr 7 mos