Umang Srivastava

• Played a key role as part of the core engineering team in architecting and delivering a Customer Care Console for enterprise clients including HDFC, WEGO, and Leocare—boosting agent productivity and transforming support service delivery. • Developed and launched high-performance Survey Data Services for major clients including Samsung and CGB, reducing survey response collection time from 10 seconds to under 1 second, significantly improving data turnaround. • Led the end-to-end integration of omnichannel messaging solutions into the Sprinklr Platform, collaborating closely with platform and infrastructure teams to improve real-time customer engagement and overall system scalability. • Spearheaded the design and development of the Agent Nudges module from scratch, integrating advanced notification capabilities to enhance operational efficiency and elevate the quality of customer interactions. • Directed large-scale Knowledge Base migration projects for high-profile clients such as Breville, POS Malaysia, and Canal+, resulting in streamlined information access and improved support workflows. • Consistently resolved complex client issues proactively, reducing average resolution time to less than 2 days and ensuring uninterrupted customer service experiences.

- Developed a malicious PDF detection module by creating a dataset of PDF files that included both malicious and benign files. - Utilized appropriate feature extraction techniques and classification algorithms like Naive Bayes, Decision Tree and Random Forest to develop a model that could accurately distinguish between them.

• Developed an innovative trading suggestion platform which recommends user to buy/sell/hold based on analytics considering technical indicators like RSI, VWAP, ATR, Moving Averages, etc. • Led the task of analyzing and formulating trading strategies with combination of technical indicators. • The platform worked on a Flask backend, with distributed servers for redundancy and fast processing. • On testing, the mathematical module gave an accuracy of 85-90%.

• Developed the official website of the organisation. Used React for frontend and Firebase for deployment. • Worked on steam heat rate optimisation for Ultratech’s power plant. Used regression and XGBoost models in H2O AI tool to find optimum steam temperature and pressure.

Worked on migration of outdated Flash based animations to browser friendly animations based on AnimeJS.