Pritesh Soni

Pritesh Soni Senior Product Developer        -about.me/priteshsoni - Bangalore, India thepriteshsoni /in/thepriteshsoni thepriteshsoni SUMMARY Backend Developer with a knack for user empathy, having 3+ years experience of working in startups. SKILLS Data Structures Algorithms System Design Problem Solving DevOps Familiarity Dialogflow Alexa ASK SDK EMPLOYMENT Active AI Senior Product Developer ‣ My role primarily involves designing and developing new features to optimize the conversational banking process and expanding the product scope to a wide spectrum of channels. I also mentor a team of 6 young developers within and beyond the business requirements. ‣ The tech stack comprises of Spring/ Hibernate, Redis, MySQL/ Oracle, ElasticSearch, and deployment automation tools like Jenkins & Docker. I am currently learning Python too. MountBlue Technologies Full Stack Developer June 2017 - May 2018 Arowana Consulting Oracle Technical Consultant Dec. 2016 - May 2017 ‣ Programming Bootcamp followed by a 9-months C2H deployment at Active AI. ‣ Worked on Oracle EBS, Alerts, Workflows, SQL, PL/SQL, and RDBMS concepts. Traveling Photography Poetry Creative Writing Bangalore Bangalore TOP CONTRIBUTIONS PDF Mining Feb. 2020 - Apr. 2020 ‣ Owned the complete backend development of a PDF Data Miner + Query Search platform. ‣ Architected and developed the PDF Miner consisting of two major processes - CSV generation from the input PDF, and consumption of CSV to extract and break down tabular and non-tabular contextual data into atomic bullet points to be saved in ElasticSearch. ‣ The Primary challenge was to extract and assort PDF content into hierarchical atomic bullet points which can be put back together to form the original context. ‣ Wrote the CSV transformer in Python and miner in Java. ‣ Achieved 85% overall data consistency and accuracy with respect to original PDF data. Multilingual support for AI Operations INTERESTS Bangalore Aug. 2017 - Present Oct. 2019 - Dec. 2019 ‣ Designed and implemented a cost-effective and optimized approach to enable conversations with the bot in multiple languages. ‣ Accelerated Export and Import operations for Smalltalk and FAQs. ‣ Implementation served as a 30% faster (in terms of TAT for a user conversation) and a 100% cheaper in-house alternative to Google Translation Services. ‣ Provided an interface to override system translated utterances for Smalltalk and FAQs. ‣ This also enabled multilingual auto-suggestions during user interactions with the bot. Product Upgrade & Release Management Apr. 2019 - Present Led Product Build and Release management lifecycle consisting of: ‣ Cherry-picking a huge array of feature implementations and bug fixes PRs from development to release branches. ‣ Conflict resolution post delta review and impact analysis, followed by submissions to QA for proper regression testing at multiple levels before cutting the release. Prioritisation Framework & Bot Analysis Jan. 2019 - Mar. 2019 ‣ The objective was to provide an advanced admin reporting module and reduce the complexity in data extraction for self-learning. ‣ Leveraged elasticsearch to audit and search every user utterance encountered by the bot in a logically categorized manner. ‣ Reduced the query processing time from 5 seconds to 0.5 seconds on the admin insights screen, consequently achieving 900% optimization! ‣ Implementation also served as a prioritization framework for clients to get a concrete report on the accuracy, overall performance of the bot and areas that needed immediate attention! Product Expansion to IoT platforms Sept. 2018 - Dec. 2018 ‣ Expanded the scope of the middleware and made conversational banking possible on IoT platforms like Google Home and Amazon Alexa and many other social channels & apps. ‣ This Omni-Channel development was instrumental in getting three new clients on board! EDUCATION APIIT SD INDIA, Staffordshire University, Panipat B.E. Honors, Software Engineering, 2017 Kendriya Vidyalaya Rajkot 10+2, Science, 2012 /