Data Scientist Learning Engineer

Manik Malhotra

+1-475-***-**** ******@***.********.*** Website GitHub LinkedIn

TECHNICAL SKILLS

Programming Skills: Python, C++, C, C#, Java, JavaScript, SQL, HTML5, CSS3, PHP Tools/Django, Frameworks: XML, Power BI, Scikit-Maven, Learn, Flask, Keras, Bootstrap, TensorFlow, Latex, AWS Pytorch, (EC2, Spark, RDS, S3, Microsoft ElasticCache) Azure, MEAN Stack, Tableau, Hadoop, Node.js, React Database Management Systems: MS SQL Server, MS Access, MySQL Version EDUCATION Control: Git, GitHub, GitLab, Agile Master University of Science of New in Haven, Data Science CT, USA Aug 2022 - GPA: Dec 2023 3.57 Bachelor Vellore Institute of Technology of Technology, in Computer Tamil Science Nadu, India Jul 2015 – May GPA: 2019 3.6 WORK EXPERIENCE

University of New Haven Feb 2023 - Present

Machine Learning Engineer

Saarthee Data Scientist • – Actionable Developing productivity Insights a and decision-improving making student and evaluation success. system for the campus's international graduate admissions office Jan 2022- for increase Aug 2022 in

• • • • Lead and Utilized means to Built emails Reported buy Created a internal the Clustering and team Machine company trends sales timeline of dashboards 7 for to analysts, in Learning the services. explore revenue of company development for Gathered and and earned, visualizing analyze Statistical by 5%business units phases . customer and modeling sold from automated requirements using interactions, dev techniques different to prod monitoring directly environment. uncover time such of dimensions from as customer KPIs, Cat Client Boost, calibrating (using behavior Comcast) Random WoW model trends, ; Forest, Translated and metrics and MoM, Time-identify that thereby into Series improved technical influencing achieving analysis, marketing roadmap factors and 23% K- CBRE Software Engineer growth. Jan 2020- Jan 2022

CBRE Associate • • • • • • Software Responsible Analyzed Handled Developed Created Provided Engineer Ad stored personal historical, Cloud Campaigns for procedures creation gates training current to such of check for websites to property as fix new Azure user runtime customers engagement on listings Key backend Vault bugs. and and for for and predicted Assisted application EMEA Derived regions churn with business security technical and through thereby and insights issues Sitecore monitor helping and for and logs. Performed Global clients SQL. regions. to retain A/B testing. 50% May customers. 2019 – Jan 2020 PROJECTS • • • • Designed Utilized Worked delivery Participated of C#collaboratively and the, HTML, in Implemented goals. code CSS, reviews and with JavaScript SQL and cross-Server contributed functional to databases create to teams, responsive continuous to store including and user improvement retrieve interfaces designers, application of and project coding implement managers, data standards server-and and side development QA logic. analysts, practices. to ensure timely Skin Spotify Finger Cancer Gesture Recommender • • • • • • • • • Detection Implemented Compared and Analyzed Leveraged system is Preprocessed Developed physical Implemented and Choose model Published only to Mouse software. Achieved users’ 3821 successfully on Manhattan contact. the out the preferences. the 6 system Spotipy a classification performance python of principle the skin the necessary results 44,best OpenCV moved cancer which 430 distance application accuracy in of submitted data Vol. is the the models detection of cv2 an as transfer cosine cursor. after 83 API package the of of based with papers. to ~83.observing distance TEST to similarity learning download classify the to 3% on Magazine code help binary with metric incorrect models cancer technique of the the random the cross-users’ backend to issued OpenCV by like calculate data entropy applying that forest personal VGG16 for format on makes library the Spyder using the loss and image month playlist which cursor's to personalized and concluded the quantify IDE classification technique a of resulted from Video to May/next enable the the they device position, June recommendations accuracy in of Spotify can communication 85% hyperparameter 2020. algorithms. to be control resulting Database accuracy. improved. of Those our the predicted results based to in between mouse tuning. model 92% on were classification pointer accuracy how a the recommendation featured camera similar without Nov May Dec after results among device a 201*-****-**** song any the