Mar 2019 – Apr 2022 · 3 yrs 1 mo · Bengaluru, Karnataka, India · Hybrid

• 1. Developed solutions for Customer Digital Experience & Marketing Campaigns, as part of R2P team with the following actions :-
• Assembled business requirements after discussions with stakeholders
• Collaborated with data engineers to identify relevant data sources and extract data
• Designed and implemented LLM-based NLP and Information Search models to enhance customer digital experiences, significantly reducing response times and improving personalization and targeting
• Designed and implemented Streamlit visualization dashboards for solution prototypes which improved stakeholder trust and adoption of solutions.
• 2. Implemented a graph-based incident visualization tool to map relationships across IT incidents for proactive mitigation of disruptions.
• 3. Developed computer vision-driven document processing pipelines for automated financial document processing, reducing manual review effort across 100+ documents and lowering downstream operational risk

Jun 2018 – Mar 2019 · 9 mos · Bangalore · On-site

• 1. Built multilingual NLP and information-extraction pipelines to process and analyze Japanese clinical trial documents, enabling insights from 30+clinical trial patient records without knowledge of the language
• 2. Designed survival analysis models to identify high-risk patients, accelerating clinical decision-making and trial efficiency
• 3. Developed data correction and augmentation pipelines for 100K+ clinical trial records,improving data quality for further analytics

Apr 2017 – May 2017 · 1 mo · Mumbai Area, India

• I worked to evaluate the pattern of demand in the airlines industry. The demand pattern could be linked with the online behavior of the consumer and other macroeconomic factors. Moreover, I worked on evaluating the effectiveness of the demand projection module used by the airlines companies.

Dec 2014 – May 2016 · 1 yr 5 mos · Kolkata Area, India · Hybrid

• 1. Developed modules to detect emotions in generic texts and add domain-specific customizations to the modules.
• 2. Developed linguistics based unsupervised methods to detect intent in text with context based customizations.
• 3. Designed and developed a NLP based search engine for the e-commerce domain.
• 4. Mentored and coached interns and was responsible for supervising 2 research associates

May 2012 – Nov 2014 · 2 yrs 6 mos · Kolkata Area, India · Hybrid

• 1. Developed Big Data based Predictive Analytics solutions for the retail and insurance industries.
• Designed and implemented solutions to coupon sorting and coupon allocation problems in the retail industry
• In the insurance industry, designrf and implementrf solution for predicting PMPM cost of patients
• 2. Developed machine learning modules that help Xpresso, Abzooba's text analytics engine to be a self-sufficient learner.
• Developed domain specific ontology using auto-generation and semi-supervised approaches.
• 3. Developed sentiment detection module for the Xpresso that was customized for 12+ industries.
• 4. Implemented data mining modules in social games analytics that can mine social games data to profile consumers based on purchases

Jul 2011 – May 2012 · 10 mos · Kharagpur

• As a Teaching Assistant for Programming and Data Structures and Database Management Systems, I used to supervise the students in their laboratory work. I was also responsible for the evaluation of the students in their projects and laboratory exams.

May 2011 – Jul 2011 · 2 mos · Kolkata Area, India

• I worked on the topic detection problem that used the linguistics based unsupervised approach to detect topics in structured text like webpages and unstructured texts like social media posts.

May 2010 – Jul 2010 · 2 mos