https://dream.weekday.works/redirectToHiringCopilotExtension/

Provider of AI-enabled supply chain management services

Platform providing real-time information & analytics that helps in understanding communities. It helps users to understand trends at the asset, collection, and market levels, and make decisions based on real-time on-chain and off-chain data. The users can explore the full spectrum of digital assets, including fungible and non-fungible tokens. It ke

Platform providing real-time information & analytics that helps in understanding communities

Developed and owned the technical vision and roadmap for foundational ML / Data platform systems at
Etsy, collaborating closely with product, engineering, and design leaders across vertical groups.

Recruited, built & led teams of 15+ engineers (and 2 managers) across San Francisco, Brooklyn, and remote locations in the ML Infra Org, delivering critical systems that enabled real-time ML solutions. 
          
Feature Data Platform- a centralized platform for managing, storing, and serving feature data to ML models.
          
Streaming Data Platform - a team that constructed datasets for features and candidate sets using
real-time data streams.
          
Realtime ML Services for Approximate Nearest Neighbors (Vector DBs) and Online Experimentation Strategies (Multi Arm Bandits Service & Interleaving)

Context Specific Ranking: Led the development of Etsy's re-ranking platform for Organic Search & Computational Advertisement ranking.

- Data engineering: Training & Feature Data extraction
- Feature Quantization & Engineering libraries
- Model training systems
- Model Deployment & Inference systems
- Launched multiple variants of search ranking models leading to major KPI wins


Developed an AI system for identifying the style of items on Etsy based on their aesthetic and visual
attributes, resulting in a paper presented at KDD'19 on "Understanding the Role of Style in Ecommerce Shopping".

Joined as part of the Blackbird acquisition. 
Continuing to work on Search, AI, Large Scale Distributed Systems and Systems Infrastructure Challenges!

-- Acquired by Etsy -- 

- Search & Ranking Infrastructure
- Diversity product ranking to provide diversity in search result set
- Spark ETL pipelines to aggregate and summarize large scale feedback data from client
- Working with pairwise, pointwise and listwise ranking algorithms to achieve state of the art e-commerce search ranking performance 
- A/B test framework to compare between different models, parameters, subset of users (live traffic)
- Ranking caching infrastructure to provide low latency dynamic ranking capabilities for search results
- Also working on Indexing stack with technologies such as Solr, RabbitMq, Scala

• Tweet suggestions for Topic Building. Latent semantic analysis over tweet documents to generate nearest neighbor word suggestions for topic building. 
• Machine Learning Insights Ranking platform. Generating insights from Twitter data - computing simulation based confidence intervals and ranking those insights based on different customizable  metrics.
• Correlation metrics for relating different sets of time series drawn over social and KPI data to best gather insights about what drives user intent.
• Wrote a Scalding - Hadoop based ETL pipeline that processes a TB of Facebook data each day. 
• Worked on our in-house in-memory Akka based realtime data system. Terabytes of tweets are kept in-memory in a bit packed format allowing low latency data slicing operations. 
• Was the lead engineer for a real time data platform project which currently powers the entire Quantifind product offering. Written in Scala with multi threading using Futures, Database layer in Postgres / Slick and web server using Unfiltered.

DNS Capital

Citi Ventures

USVP

Valor Equity Partners

S&P Global

In-Q-Tel

Snowflake

Redpoint Ventures

AME Cloud Ventures

IRIS

Comcast Ventures

Cathay Innovation

Andreessen Horowitz

National Science Foundation

Plug and Play Tech Center

Archer Venture Capital

Sand Hill East

Quantifind provides AI-based brands engagement and financial crime intelligence solutions. Offers a SaaS-based brand engagement solution for generating insights from market data. Allows companies to benchmark against specific competitors or an entire category. Offers a fraud and AML intelligence solution, that automatically assess risk factors for individuals and organizations. It automates the search for entities and their relationships across data sources for fraud including social media and deep web forums.

AI-based brands engagement and financial crime intelligence solutions

Worked with Professor Jaime Carbonell on unsupervised Machine Learning problems around matching skills with candidates in the field of recruiting.

National Security Agency

Yahoo

Carnegie Mellon University founder Andrew Carnegie said: "My heart is in the work."

No statement better captures the passion and drive of our people to make a real difference. 

At Carnegie Mellon, we're not afraid of the work.

Our educational environment creates problem solvers, drivers of innovation and pioneers in technology and the arts.  

Employers in every field say our graduates are ready to hit the ground running the day they graduate. 

So, join us. Whether you're looking for a career or an education. Or both.

University offers multiple courses across engineering and business mangement

Teaching Assistant for Introduction to Computer Systems
https://www.cs.cmu.edu/~213/

•	Investigated open source parsing (OpenCSV) and rewrote the entire Hadoop Job for Data Ingest in order to make it 2x-3x faster. 
•	Wrote an Amazon Elastic Map Reduce Emulator to build lenses (in-memory summary of Hadoop data) locally over an EMR environment.
•	Improved usability around getting appropriate Hadoop counters, warnings and pulling job configuration from MR jobs.

Platfora is an end-to-end software platform to run natively on Hadoop and spark, to provide raw data preparation, in-memory acceleration, and rich visualizations to better share insights.  An interactive and visual full-stack platform delivered as subscription software in the cloud or on-premises, Platfora Big Data Analytics is creating data-d

• Designed and implemented a real time video trending pipeline in Scala / Storm (framework).
• The final version given raw player logs could compute top k videos based on 5 different
algorithms and then these results could be filtered with various key combinations.
• Deployed and set up the staging (testing) environment by writing a Hadoop Job to continuously
feed raw player logs and extract useful information from them.