Heet Shah

Machine Learning Engineer

Ahmedabad, Gujarat, India6 yrs 10 mos experience
Most Likely To SwitchAI Enabled

Key Highlights

  • Expert in Machine Learning and AI solutions.
  • Proven track record in cost optimization.
  • Strong background in recommendation systems.
Stackforce AI infers this person is a Machine Learning Engineer specializing in SaaS applications with a focus on AI-driven personalization.

Contact

Skills

Core Skills

Machine LearningRecommender SystemsArtificial Intelligence (ai)

Other Skills

AirflowCI/CD PipelineCollaborative FilteringComputer VisionData ScienceData StructuresDatabricks ProductsDocker ProductsDocumentationFeature EngineeringFeature SelectionFlaskGenerative AIGoogle Cloud Platform (GCP)Google Cloud Platform services

About

"Great minds discuss ideas. Average minds discuss events. Small minds discuss people." - Eleanor Roosevelt Hello - I'm Heet, working as a Machine Learning Engineer on various problems at large scales of millions of users/items. Working on ML techniques for Recommendation, and personalization, Content and User understanding, Trust & Safety: Moderation AI and Feed AI: Multi-objective decisioning. Passionate about AI - Machine Learning and Deep Learning, Recommendation Systems, Pattern Recognition, Problem Solving, Software Optimization. With having profound interest and experience, some of the key skills and concepts that I am interested in: - AI: Recommendations System, Computer vision, Audio/Speech processing, NLP - Problem-Solving skill: Data Structure and Algorithms - Backend development and optimization: MicroService, Job deployment - SQL/NoSQL Databases - Google Cloud Platform services - CI/CD Pipeline: Docker, Jenkins, Kubernetes - Flask Want to explore things in-depth and make an impact in society by uplifting people's lives along with personal advancement as well.

Experience

Sharechat

2 roles

Senior Machine Learning Engineer

Promoted

Apr 2023Present · 2 yrs 11 mos

  • Worked on User / Item personalization problem, specifically the Candidate generation layer for serving millions of users.
  • Feed AI:
  • Recall models: Matrix factorization, Content Embedding based Candidate Generator (CG), User state model-based CG (partial contribution)
  • Topic-coherent CG: Based on an interacted topic by the user, suggest similar videos to capture the user interest. Similar posts were recommended by learned topic embedding for a given post.
  • Recent-aware CG: In-session personalization kind of CG, where it would account for user's last k interaction and based on that similar items would be suggested. Similarity is measured by a real-time learned vector representation of user and item based on their interaction.
  • Two-tower CG: Incorporate different user and post features into CG to capture users personalization better on short-term as well as long-term aspects.
  • Cost Optimization::
  • ~20M+ INR / quarter savings were observed by doing various optimizations at job/service - deployment level in Content Understanding pipelines.
  • [Reasons]: By reducing the data transfer cost from one region to another, setting up proper HPA metric, CDN cache fill / CDN lookup cost savings, Using GCS bucket client API instead of using public storage url, using spot compute instance instead of reserved instance
Python (Programming Language)KerasModelingDocumentationModel TrainingCollaborative Filtering+18

Data Scientist

Aug 2021Apr 2023 · 1 yr 8 mos

  • Working on ML challenges around Content and User representation, Trust & Safty of platform: Moderation AI and Feed AI: personalization and recommendation for Moj, ShareChat & MX Takatak (350M+ Monthly active users).
  • Have worked on several projects, here are few of the key contributions:
  • Content and User representation:
  • Face clustering algorithm
  • Gender/Age/count-of-people detection
  • Continual/Online learning for Topic prediction
  • Indic Language Identification
  • These representation can ultimately be useful in Feed AI for personalized recommendation and user retention.
  • Moderation AI:
  • NSFW detection models/Improvements in the models for ShareChat, Moj, Livestream (currently 1M+ daily active users)
  • NSFW video Explainability model for finer annotation
  • Improved model has an Impact of cost saving of >4M+ INR per quarter in terms of Moderation cost.
  • Analyzing content came on the platform Sharechat and Moj as a part of the Content and User Understanding team, wherewith the help of ML/DL and statistics will try to get insides of content/user which can eventually help in various User retention aspects.
Python (Programming Language)KerasModelingDocumentationModel TrainingFeature Selection+16

Crest data systems

Software Development Intern

Jan 2021Jun 2021 · 5 mos

  • Project is to Migrate ElasticSearch-Logstash-Kibana (ELK) server data to the Splunk server and automate the migration process with just one click. To enhance one of the production tools to satisfy the criteria and enhance the features.
  • Worked on ElasticSearch Connectors to introduce 3rd party platforms for the alert generation.
Python (Programming Language)DocumentationSoftware SystemsTeamwork

Infivolve, inc.

Machine Learning Intern

May 2020Sep 2020 · 4 mos

  • Project was to construct a reliable virtual Exercise Trainer which can classify different 100+ exercises like pushups, plank, high jump, shoulder press etc in Real time, and can able to give proper feedback to correct the exercise, in case if any person doing it wrongly.
  • We have used Dynamic Time Wrapping algorithm to measure distance from an ideal exercise. Exercise trainer can also able to justify about dominant part of body (Left/Right), by detecting and tracking movements of body keypoints.
  • Need to construct Flask based web application for exercise testing and analysis purpose using HTML, CSS, Javascript, and python
Python (Programming Language)ModelingModel TrainingFlaskArtificial Intelligence (AI)Software Systems+3

Scaler academy

Campus lead: Scaler Achiever by InterviewBit - [NonTech]

Mar 2020Oct 2020 · 7 mos · India

  • Scaler Academy has one goal to make a user-friendly environment for coding on different campuses. Being the part of the scaler achiever club, the responsibility is to make awareness of the activity driven by the finest product of an InterviewBit. Coding competitions awareness in the campus and organization and promote the events based on the Scaler club.
TeamworkLeadershipcollaboration

Meditab software india pvt. ltd. (dosepack india llp.)

AI Research Intern

Dec 2019May 2020 · 5 mos

  • The Project involves an Automation of seal packing of the pills, prescribed by the Doctors. It includes the building of a Machine learning prototype that can identify pills-different types of tabs inside the box, how many pills are there, classification, and masking of those pills. The tedious task is to form a Generalized Feature Extractor that can adapt new classes Dynamically at run time.
ModelingDocumentationModel TrainingFeature SelectionArtificial Intelligence (AI)Software Systems+3

Computer society of india

2 roles

Public Relation Officer (Core Member) - [NonTech]

Nov 2019Oct 2020 · 11 mos

Committee Member - [NonTech]

Jun 2018Oct 2019 · 1 yr 4 mos

Education

Nirma University

Bachelor of Technology — Computer Science and Engineering

Stackforce found 100+ more professionals with Machine Learning & Recommender Systems

Explore similar profiles based on matching skills and experience