Pranjal Pathak.
Name: Pranjal Pathak
Profile: Data scientist
With a broad range of experience implementing machine learning across domains and strategic solutions in Natural Language Processing, I have bootstrapped four end-to-end projects from design to production. I have gained recognition and a handful of awards for my contributions during my aggressive and demanding roles.
My first campus recruitment saw me recruited as an early-age ML Researcher and I have been steadily working towards competitive positions ever since. From being the youngest Technical Lead developing NLP products to winning the Leadership Award amongst 14 other VP-level members at Morgan Stanley’s highest global ceremonial event - I have come a long way.
I have always been given a free-hand at driving innovation tracks and have successfully converted various POCs to realization.
3400
LINES OF PYTHON6
YEARS OF EXPERIENCE16
TOTAL PROJECTS5
AWARD WONLed multiple award-winning innovation tracks and development of global Chatbot engine, IE extraction tools.
Boostrapped PDF Meta-data extraction Engine and real-time email-attachment classifier, reducing manual intervention by 75% for Zycus's leading AI Suite.
Researched and developed three time-series forecast IPs for Wipro's innovation ML lab implementing 12 successful client conversions.
Deep-diving into Deep Learning, Advanced NLP, Statistics and Data Mining techniques to be ahead-of-the curve industry-oriented Data Scientist.
Learned state-of-the-art deep learning model development & analyzing semantics in Natural language for I.E engines.
Learned Engineering principles, applied mathematics & smart Robotics. Discovered my love for data-driven analytics using computer science.
Explored the world of Science and acheived academic brilliance in diverse learnings.
“Life is what happens when you're making other plans. A dead end is really a detour to a new route you hadn't planned on taking. Every experience enhances your life, so does your Résumé.”
These are some of the projects that I have worked on and are open-sourced.
This projects utilizes Open-AI's open-sourced NLG components like Da-Vinci Completion and Codex to fit in the user-prompted Information Extraction pipeline.
Analyzed anxiety and sentiments, behavioural mood patterns across pyschiatrist-family-patient conversational transcripts for helping CMI hospital deduce text analytics.
Extracted meta-data such as Company Name, Address Block, Phone number from scanned financial PDF invoices using supervised learning approach and two-layer contextual classifier engine.
Using negative-sampling and IR score ranking DNN mechanism, approached SQUAD dataset for extracting ranked list of candidate documents and potential paragraphs list.
I occasionally take on freelance and mentoring opportunities too. If you have an exciting project where you need some help, or wish to know more about my work experience - Send me over a message, and let's chat.