Janak Sachdeva Email & Phone Number

Bengaluru, Karnataka, India

Working on Import Export functionality of Workflow Templates.Working on building Workflow Templates functionality for Intelligent Agreement Management.

Bengaluru, Karnataka, India

Building Intelligent Agreement Management Platform.

Noida, Uttar Pradesh, India

Worked as a part of Reliability crew in PPTO improving crash rate of the app by 50% & had set up alerting mechanism in case of crash spikes. Also fixed few other errors which led to non dismissible dialog in the app.Worked as a part of Accessibility crew: Implemented New Thumbnail Pane in PPTO which supports Accessibility and has Sections support by reusing the existing Slide Sorter code. Also fixed issues in Slide Sorter view on the fly so as to avoid them in New Thumbnail Pane.Worked as a part of PPTO Shield crew: Fixed issues in the product from various domains like Slide Sorter, Accessibility, Pictures, Ink etc. to improve user experience.Implemented Shape Insertion of variable size at desired location by mouse drag in PowerPoint Web.

Noida, Uttar Pradesh, India

Worked on improving the boot performance of PowerPoint Web app:• Responsible for creating various PowerBi dashboards for monitoring boot performance related metrics like TTI and TTVR.• Worked on identifying and preloading boot critical files in browser, so as to improve boot time of the app.• Worked on automating boot performance related alerts and creating IcMs so as to capture any rise in boot perf. metrics. Upon increase in any boot perf metrics, an alert in the form of email and IcM gets generated. Improved TSGs for these alerts.• Improving TTVR for some buttons like Sensitivity, Dictate on ribbon present in PowerPoint Web app, so as to improve user experience on booting. Used A/B testing to check for gains in sensitivity button's TTVR improvement.• Decreased running time of aggregation jobs which populate boot related metrics to one third of existing time and improved their reliability. • Added a boot metric to experimental scorecard.• Fixed some issues related to Slide Sorter View etc. and worked on dead code cleanup which was no longer being used.

Bengaluru, Karnataka, India

• Applied Topic Modelling techniques over Eat.fit customer reviews: Clustering similar reviews together for analysis.• Used Word Embeddings based approach for clustering similar customer issues together for eat.fit in unsupervised manner.• Implemented a classifier using word embeddings and neural networks for Level 1 (having 9 labels) classification of emails received over curefit’s customer support mail (accuracy:89%). Did the same thing for Level 2(having 26 labels) and achieved 74% accuracy (an increment of 6% over existing Bernoulli naïve Bayes).• Technologies Used: Python, Gensim, GloVe, Keras, nltk etc.

Bachelor Of Technology - Btech, Information Technology, Cgpa : 9.27/10

Dept Rank: 9/128

Senior Secondary School Examination, 91.6%

Secondary School Examination, 10 Cgpa