Work
Avalara
Avalara is a tax technology company. It bought Indix in 2019. I led the teams that worked on product classification and tax content sourcing.
Product classification puts a product into the right tax code, so the correct tax rules apply to it. My teams also classified products into Harmonized System codes for cross border shipments. Tax content sourcing crawls the rules and rates that tax authorities publish and parses them with NLP, so the tax engine stays current.
Indix
Indix set out to build the world’s largest product catalog. It crawled product pages from more than 5,000 brand and retailer sites. It classified products into a taxonomy of more than 5,000 nodes. It extracted product attributes and matched the same product across retailers. Customers used the data through a search API.
I first worked on the crawler, built with Scala and Akka. Then I led the team that built the data pipelines on the Lambda Architecture. Later my team worked on NLP problems such as product classification and attribute extraction. In 2018 we built an e-commerce knowledge graph with 100 million nodes and about a billion edges, used for query understanding in product search. At the time of the acquisition I led the machine learning and data platform teams.
Posts from this time are in the Indix Archive.
ThoughtWorks
I was a technical lead on GoCD, a continuous delivery and release management product. ThoughtWorks later made GoCD open source. Before GoCD I was a technical lead on client projects in e-commerce and social media.
Indigo Architects
Indigo Architects was a travel product startup. I worked across engineering, product and people roles.
Talks
- Lessons Learnt Building Domain Specific NLP Pipelines. Scale By the Bay, San Francisco, 2019. Slides.
- Continuous Delivery Principles for Machine Learning. Scale By the Bay, San Francisco, 2017.
- Modern Software Architectures and Data Pipelines, panel. Scala By the Bay, San Francisco, 2016.
- Analytics on Large Scale, Unstructured, Dynamic Data using Lambda Architecture. The Fifth Elephant, Hasgeek, Bangalore, 2014. Slides.
- Slides for the Pune Scala meetup, 2014.
Interviews
- Deep Learning has taken Natural Language Processing by storm. Written interview, Scale By the Bay, 2019. The original page is offline, so this links the Internet Archive copy.
- Interview at Scale By the Bay, video, 2017.
- Interview at Scala By the Bay, video, 2016.
- At The Fifth Elephant 2024, Hasgeek, video. I came back 10 years after my first talk there.
Education
Bachelor’s degree in Computer Science from Pune University, with a gold medal. In the final year my team represented India at the Asia finals of the Microsoft Imagine Cup.
More code is on GitHub.