Hello World!

I'm Dylan Saldanha

One of the top 10 hires by LTI in 2019, selected through Infinity Coding Challenge conducted all over India with 5000+ applicants. A 10x programmer with 2.5+ years of experience in building solutions for Intelligent Automation using Python, Javascript, NLP, Chatbots, SQL, and prebuilt tools, and passionate about building scalable and resilient solutions which can automate/expedite business processes. Currently, working on a solution for auto creation of knowledge graphs from resumes in order to facilitate the process of filtering them, and finding the right candidate.

img01

General Info

  • Date of BirthOct 13, 1996
  • LocationBangalore, India
  • Current OccupationSenior Software Engineer at LTI
  • LanguagesEnglish, Hindi, Kannada and other local languages
  • E-maildylaniansld@gmail.com
  • Phone+91 9480221970

Experience

  • Aug 2019 - Present

    Senior Software Engineer

    Larsen & Toubro Infotech

    • Mainly worked on solution building, architecture design, client proposals, PoCs and tool competency building, as a core contributor for the Intelligent Automation CoE.
    • Architected and developed the backend for an end to end Policy Review solution, using Angular, Python, Flask, MySQL, AWS Textract & Azure Form Recognizer, which assisted brokers in reviewing policies and binders, in turn reducing the AHT by 45% - 60%.
    • Designed and lead the development of a Digital KYC Agent , using Dialogflow, Cloud Vision, Python and Flask on GCP, to extract data from Aadhaar and PAN, resulting in 80% TAT reduction in the customer on boarding process.
    • Designed and developed a Digital Inventory Assistant, using Dialogflow, Cloud Vision, Python, Flask and Firestore on GCP, to help with inventory related queries, item availability check, ordering and tracking.
    • Lead a team of 5, in building the Accelerator Marketplace, to buy/sell reusable pieces of code across internal departments, reward contributors with gift cards and allow users to test code on platform; built using Python, Flask, Angular, MongoDB and Azure Web Apps.
    • Designed a custom workload manager, using Python, to adhere to SLAs and maximize Automation Anywhere bot license utilization.
  • Jan 2019 - June 2019

    Data Science Intern

    Nuveda Learning

    • Built a domain-based spam document classifier using CNN and Doc2Vec (trained on Google News dataset) with an accuracy of 94%.
    • Built an ALSTM to predict if the user will post his/her learning on a particular day with an accuracy of 92%.
    • Deployed the above models on to AWS Sagemaker and used AWS Lambda to allow API access.

Education

My Projects

  1. Hydrocarbon Exploration using Seismic Imaging

    Jan 2019 - May 2019

    Finding hydrocarbons through salt detection has been an intricate process, ever since they images salt the first time. Our solution to the problem is to basically use U-Nets which have shown state of the art results on image segmentation. Each pixel in the image is checked for the presence of salt and further on this is how the proportion of salt in that seismic image is calculated. Our model boasts an accuracy of 94.7% which is significantly higher than that of in the previous implementations.

  2. WeByte

    July 2018 - August 2018

    WeByte is a programming colosseum that allows programmers to battle against other programmers in competitive programming contests. This Flask based website can be used by an individual who wishes to host a competitive coding event in one's locality, school, college or university. Companies can also host this website on their servers and use it for competitive coding tests for hiring purposes, instead of using a middleman to conduct the tests.

  3. Analysis of Stock Prices

    May 2018 - June 2018

    Analysis of Stock Prices of a set of American Banks between 2006 and 2016.

    Focuses on exploratory data analysis of stock prices. Dataset retrieved from Google finance. Disclaimer: This project is not meant to be a robust financial analysis or be taken as financial advice.

  4. Analysis of 911 Calls

    April 2018 - May 2018

    Analysis of 911 Calls made between 2015 and 2016.

    Analyzing 911 call data from the year 2015 and 2016. Dataset found on Kaggle.

  5. Simulation of Dijkstra's Algorithm

    April 2018 - May 2018

    A simulation of Dijkstra's Algorithm using C++ and OpenGL.

  6. MyStory

    February 2018 - Present

    A Flask based website that allows users to build their own time lines in famous literary works including plays, novels, movie scripts etc.

  7. Analysis of Sales and Purchase

    February 2018 - July 2018

    A Maven based software to keep track of all sales and purchase activity in a company. This was built as a product for 2 companies that requested it.

  8. Attendance Management System

    Novemeber 2017 - December 2017

    Attendance Management System deals with the maintenance of the student’s attendance details. It engenders the attendance of the student on the substratum of one's presence in class. It is maintained on a circadian substructure by the faculty members of the university.

  9. Feedback System

    April 2017 - May 2017

    A College System to get feedback from the students about the faculty members teaching them. Mainly consists of four types of users namely, student, faculty, head of department and admin.

Skills

"The mark of higher education isn't the knowledge you accumulate in your head. It's the skills you gain about how to learn." —Adam Grant

  • Utilise interpersonal skills to motivate and encourage co-workers, understanding that major goals are achieved through teamwork.

  • Charismatic and confident public speaker.

  • Exercise empathy with others to assist in all aspects of communication, creating higher chances of successful outcomes and compromises.

  • Identify efficient solutions for given tasks and successfully implement them, aiding in effective problem-solving.

  • Tech savvy, with the ability to learn and use new methods, policies and technological advancements as they arise.

img04

Specialisation & Certification

img05

Hobbies and Interests

  • Hobbies :

    • Participating in programming competitions, usually on HackerEarth (acquired a rank of 63 in Infinity Code Wars 2018 conducted by Indian Institute of Science Bangalore).

    • Been playing the piano since the age of 6, currently a Grade 4 pianist and backup singer in a band.

    • Swimming, to stay fit and healthy.

    • Cooking, to quench that inner foodie’s thirst.

  • Interests :

    • Data Science, as it facilitates individuals to make better informed decisions.

    • Machine Learning, due to which machines can self learn without being explicitly programmed.

    • Deep Learning, trying to simulate a human brain would fascinate anyone.

Goals

"People with goals succeed because they know where they're going."—Earl Nightingale

  • To publish a book within the next 2 years on how anyone can learn Deep Learning.

  • Become a Data Scientist within the next 5 to 6 years and try to tackle real world problems like rare diseases et cetera.

  • To bring about a change in the Indian Education System so as to boost the quality of personnel from India.