Author Image

Hi, I am Dalon Lobo

Dalon Lobo

Student at St. Francis Xavier University, Canda

I am a passionate data scientist professional with 7 years of work experience who loves to build machine learning models. Currently, studying Post baccalaureate in Artificial Intelligence at StFX University, Canada. My vision is to build a startup which can create jobs for many. I work on some fun projects such as machine translation, computer vision, etc. I am open to work, do checkout my profile and get in touch with me.

Data Science
Developer
NLP
AI/ML
Team Work
Hard Working

Skills

Education

Post-Baccalaureate Diploma in Artificial Intelligence
GPA: 94.82 out of 99
Taken Courses
Course NameTotal CreditObtained Credit
Machine Learning9999
Evolutionary Computation9998
Coding in Health Analytics9999
Programming & Data Structures9998
Computer Organization9994
Extracurricular Activities
  • Website developer for Antigonish School Supplies Share Program
  • Volunteer for StFX Students' Union - Kevin's Food Resource Center
B.E. in Electronics and Communication Engineering

Experiences

1
Videoken Software Pvt. Ltd.

Dec 2017 - Dec 2020, Bangalore, India

VideoKen offers AI based video product solutions to turn videos into interactive & immersive learning experiences.

Consulting ML Engineer

Sep 2020 - Dec 2020

  • Expanding and optimizing search and Indexing algorithms
  • Monitoring and maintaining key APIs/services
Research Engineer

Mar 2018 - Sep 2020

  • Designed and implemented an end-to-end search engine using elastic search and BERT embedding to achieve semantic understanding of the search terms, which drives the sale of Video lake. This includes dockerized microservice for RESTful backend and model inference
  • Designed and implemented an end-to-end Acoustic Speech Recognition(ASR) pipeline, using the Baidu’s DeepSpeech2 ASR architecture trained on 1000 hrs of custom transcribed audio and reduced the Word Error Rate to 8%, which helped Videoken’s AI based video platform to increase video engagement by 30% in one week. This includes RESTful backend with Flask, Cosmosdb, and model inference on GPU machine
  • Implemented punctuation pipeline which restored missing inter-word punctuation for video subtitle (SRT) files using bidirectional RNN with attention mechanism (punctuator2) architecture
  • Designed and implemented video recommendation, using the sentence embeddings of the metadata of the videos
  • Implemented Speaker Diarization pipeline by training a Fully Supervised Speaker Diarization (UIS-RNN) model on 50GB of audio-text data which helped in implementing the Sentence Segmentation algorithm for indexing the videos
  • Written technical design specifications which clearly states the algorithm and implementation details
  • Served as a core group member in defining and prioritizing technology investments and ensuring the alignment of process, technology and business objectives
Research Engineer Intern

Dec 2017 - Mar 2018

  • Research on AI architecture and pipeline enhancements to Video Indexing

Software Developer
Freelancer

Mar 2017 - Dec 2017, Bangalore, India

Responsibilities:
  • Designed and developed a website for Pied Piper Events pvt. ltd., which includes dynamic pages, chat, blog and admin interface using Django-Python and hosted on AWS server
2

3
Software Engineer
Accenture Pvt. Ltd.

Sep 2015 - Feb 2017, Bangalore, India

Multinational professional services company that specializes in information technology services and consulting.

Responsibilities:
  • Designed and developed a dashboard for Credit Suisse Bank using Django-python, which reduced the report generation time by 50%. This project includes SQL integration and hosting on Nginx
  • Trained the internal team on Python for automation

Software Developer
Telenetix Pvt. Ltd.

Jun 2013 - Jul 2015, Manipal, India

Telenetix provides cost-effective technology for contact centers, all aimed to streamline business processes and optimize customer care.

Responsibilities:
  • Designed, developed and deployed end-to-end solution with dashboard using Django framework, which is used for remote monitoring and controlling the Solar Power Plant in Australia
  • Designed and developed Tx-Contact monitoring dashboard, which is used by Contact Center supervisors to generate reports and monitor agents
4

Projects

Acoustic Speech Recognition
Research and Developer

Designed and implemented an end-to-end Acoustic Speech Recognition(ASR) pipeline, using the Baidu’s DeepSpeech2 ASR architecture trained on 1000 hrs of custom transcribed audio and reduced the Word Error Rate to 8%, which helped Videoken’s AI based video platform to increase video engagement by 30% in one week. This includes RESTful backend with Flask hosted on nginx(AWS), Cosmosdb, and model inference on GPU machine. Built a preprocessing pipeline, which downloaded the video, split the converted audio before speech to text, post-process the text into SRT (subtitle) file which syncs with the original video.

Search Engine using Elastic Search
Research and Developer

Designed and implemented an end-to-end search engine using elastic search and BERT embedding to achieve semantic understanding of the search terms, which drives the sale of Video lake. This includes dockerized microservices for RESTful backend and model inference.

Inter-word Punctuation Restoration
Research and Developer

Implemented punctuation pipeline which restored missing inter-word punctuation for video subtitle (SRT) files using bidirectional RNN with attention mechanism (punctuator2) architecture.

Video Recomendation using Sentence Embeddings
Research and Developer

Designed and implemented video recommendation, using the sentence embeddings of the metadata of the videos.

Speaker Diarization
Research and Developer

Implemented Speaker Diarization pipeline by training a Fully Supervised Speaker Diarization (UIS-RNN) model on 50GB of audio-text data which helped in implementing the Sentence Segmentation algorithm for indexing the videos.

Full Stack website development
Software Developer

Designed and developed a website for Pied Piper Events pvt. ltd. (https://piedpiperevents.com), which includes dynamic pages, chat, blog and admin interface using Django-Python and hosted on AWS server.

Dashboard for Reporting
Software Developer

Designed and developed a dashboard for Credit Suisse Bank using Django-python, which reduced the report generation time by 50%. This project includes SQL integration and hosting on Ngnix.

Remote Solar Power Plant control and reporting
Software Developer

Designed, developed and deployed end-to-end solution with dashboard using Django framework, which is used for remote monitoring and controlling the Solar Power Plant in Australia.

TxContact Dashboard
Software Developer

Designed and developed Tx-Contact monitoring dashboard, which is used by Contact Center supervisors to generate reports and monitor agents.

Facial Expression Detector and English to Kannada Translator
Machine Learning Engineer

Visualize and explore FER-2013 dataset with 28k images, implement CNN architecture with prediction on webcam images converted to emoji. English to Kannada Translation using Bilingual Sentence Pairs dataset using RNN, GRU and LSTM models.

An Exploration of Different Techniques for Artificially Producing Art
Software Engineer

An Exploration of Different Techniques for Artificially Producing Art using Genetic Algorithm.

Accomplishments

Post Graduate Program in Big Data and Machine Learning

In this course I learnt to work with Big data using Spark and Hadoop. Learnt basics of Machine Learning, explored data visualizations, feature/model selection, tuning and introduction to analytics using statistical techniques. Click here to view my projects.

Python for Data Science and Machine Learning Bootcamp

This course taught me how to use NumPy, Pandas, Seaborn, Matplotlib, Plotly, Scikit-Learn, Machine Learning, Tensorflow.

This course gave me a broad introduction to machine learning, datamining, and statistical pattern recognition. Topics included: (i) Supervised learning (parametric/non-parametric algorithms, support vector machines, kernels, neural networks). (ii) Unsupervised learning (clustering, dimensionality reduction, recommender systems, deep learning). (iii) Best practices in machine learning (bias/variance theory; innovation process in machine learning and AI).

Data Science A-Z™: Real-Life Data Science Exercises Included

This course taught me Data Science step by step through real Analytics examples. Data Mining, Modeling, Tableau Visualization.

Machine Learning A-Z™: Hands-On Python & R In Data Science

This course taught me hands on machine learning using python and R programming languages.

Introduction to R

I learned the basics of data analysis by manipulating common data structures such as vectors, matrices, and data frames using R.