Aashish Dattani

Software Engineer

New York City, New York, United States13 yrs 11 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Stackforce AI infers this person is a SaaS and Healthcare expert with strong skills in Machine Learning and Search Technologies.

Machine LearningRanking AlgorithmsData StructuresSearch AlgorithmsInfrastructure EngineeringBioinformatics

AerospikeApache SolrCC#C++Data MiningHTMLJavaLinuxNoSQLPythonSQLSearch AnalyticsSolrXML

Apr 2022 – Present · 3 yrs 11 mos · New York, United States

1. Ranking engineer in the Google Lens Multimodal Ranking team: Part of the effort to make the Google web search ranking stack to serve multimodal queries containing images. Worked on core ranking components of the Google Search stack, and evaluated improvements through offline and live experiments.
2. Google Indexing engine: Worked on adapting Google's crawling, processing and indexing engine to work with Human Voices data (including social media content), which have a high publish rate and freshness requirement. Increased the freshness of these contents by 50% in the Google index.
3. Google Videos ranking: Worked as ranking engineer in the Video ranking team, helping index and rank short video content (from platforms such as Tiktok, Instagram etc) on the main Google search results page. Helped improve clicks and conversions by +10% for these contents.

C++CJavaPythonMachine LearningRanking Algorithms

Nov 2016 – Apr 2022 · 5 yrs 5 mos · San Francisco Bay Area

1. Apache Solr Engineering: Implemented components for implementing custom search logic: ranking function, elevating and slotting documents and customizing the TF-IDF similarity logic. The custom components led to an increase in search revenue of 15% at launch.
2. Search Analytics Pipeline: Built a framework for search analytics from JS clickstream data; attributing conversions to search queries, and identifying worst-performing queries for feedback to search algorithms. This is the critical training data for all ML frameworks used for search.
3. Search Relevancy Framework: Designed a framework for combining various relevancy signals obtained from ML classifiers. This involved normalizing all input signals to be combined into a single weighted-linear combination function that is used for scoring in Solr.
4. Classifier Training Optimization: Optimized the training pipeline used by search relevancy classifiers. 30% reduction in training time achieved by pooling training instances into a single vector and using training probabilities to compute softmax loss function.

Apache SolrMachine LearningSearch AnalyticsJavaPythonSearch Algorithms

Apr 2014 – Nov 2016 · 2 yrs 7 mos · Bengaluru Area, India

1. Early Engineering Team: Part of the that built an e-Commrce search-as-a-service SaaS company, that was later acquired by Target Inc.
2. Solr Infrastructure Setup: Built and scaled the entire multi-tenant Solr infrastructure to handle millions of catalog updates and guarantee customer SLAs for enterprise clients.
3. No-SQL Aerospike Database: Deployed and scaled a NoSQL database with replication and fault-tolerance to handle real-time catalog updates while guaranteeing 99.9% uptime.

SolrNoSQLAerospikeJavaInfrastructure Engineering

Apr 2013 – May 2013 · 1 mo · Bangalore

Served an IT services client in improving operational efficiency by designing a staffing and supply chain strategy.

Jul 2010 – Jun 2012 · 1 yr 11 mos · Bangalore

Developed statistical algorithms and UX designs for analyzing next generation DNA-sequencing data as part of a product team in a bio-informatics startup. Used Java, Python and C for implementing analytical routines to efficiently handle large genomic datasets.
Conceptualized and implemented statistical tools for identifying large-scale cancer-causing DNA variations using "big data" analytics on population genomic samples.

JavaPythonCBioinformatics