Jan 2023 – Mar 2024 · 1 yr 2 mos

May 2022 – Apr 2023 · 11 mos

• Drishti has now been acquired by Apple.
• The organization covered the hardest problem in Computer vision - Action detection, in a very traditional entrenched setup.
• Led the Engineering team towards defining and implementing the product and engineering roadmap, including critical initiatives such as on-premise and cloud agnostic changes, self-serve AI, newer AI models for action detection, and enterprise readiness.
• Multiple ML approaches for object and action detection were tried and deployed - from 2D convolution with LSTM, 3D convolution, I3D, and BMN to transformers based DETR + LSTM.
• Led the cloud-agnostic version of Drishti, porting it to various public clouds and VMware on-premise stack. Coded much of the initial terraform for public port and early POCs on VMware stack.
• Grew the team to 35 members, including later leading scaledown and strategic shifts. Built and implemented a detailed Engineering evaluation rubric.
• The engineering structure brought in was much more scalable, empowered, efficient, and aligned with the organization's and engineering goals.

Jun 2021 – Apr 2022 · 10 mos

Aug 2020 – May 2021 · 9 mos

• Building an application platform for the manufacturing industry using AI, computer vision & scalable systems.

Mar 2020 – May 2020 · 2 mos

• Leading technology & AI initiatives against Covid-19 crisis as a short term consultant. Interventions were designed as pilots or prototypes:
• Designed & prototyped the self-screening chatbot that influenced similar functionality in Aarogya setu. Worked on initial designs of an extended contact tracing app.
• Led and built WhatsApp & IVR based tools for local Covid-19 related citizen surveys, deployed in some cities.
• With researchers from IISc, St John hospital and others, helped design and draft the AI-based drug repositioning proposal.
• The Unlockdown project has built protocol and technology for enabling safe workplaces to reopen post the lockdown.
• BillionSocialMasks.org helps to provide the much-needed quality affordable masks and personal protective equipment. Part of the core team.

Apr 2016 – Jul 2020 · 4 yrs 3 mos · India

• Took on the problem to modernize the fragmented $100B lifestyle market, dominated by more than 4 million small businesses. Conceived and built Wishbook app, India’s first fashion focussed B2B marketplace and SaaS platform:
• Led the product and engineering spread across 2 offices, building the highly scalable marketplace and SaaS platform on various open source stacks (Django, MySQL, ReactNative, Android, AngularJS). The engineering followed feature scalable architecture - with sprints, wireframes, clean interfaces, Rest APIs, automation testing.
• Capital efficiently built and scaled the company, product, systems and marketing to a team of 60+ employees, with the marketplace profitably doing an annualised USD 2M+ GMV.
• Recognised for its unique approach to solving a complex fragmented at multiple forums and awards, including Best B2B eCommerce (Franchise India, 2018) and Most Promising Fashion Startup (Images Fashion Awards, 2019)
• Raised seed and pre-Series investments for a from marquee investors.

Jan 2012 – Mar 2016 · 4 yrs 2 mos · Surat Area, India

• Triveni sarees is a traditional family held apparel manufacturer with 150+ employees and turnover of $15M. Joined to upgrade systems, processes and positioning:
• Took the brand online in a bootstrapped sustainable profitable manner through a multi-channel sales model. Created website and app, completely integrated multi-channel sales and order management software. The entire business unit built from scratch grew to a team of 50+ members across all functions, including software engineers.
• Developed, built and product lifecycle management and warehouse inventory systems and processes, deploying with the existing team of 20 data entry operators.
• Digitised and implemented the field and channel sales force and field sales processes for better results, changing behaviors of more than 100 field sales executives.
• These interventions saw Triveni take leadership in efficient operations and digital presence over its competition.

Jan 2008 – Jan 2011 · 3 yrs · Mumbai Area, India

• Took on the mandate to solve one of the most pressing problems in India - providing affordable quality healthcare to the lowest income segments of population, targeting 15 million people by 2015:
• Developed core software systems and MIS, which was one of the first fully cloud and browser based health management systems covering preventive, primary, tertiary care, including insurance processes and financing. Deployed through a set of 4 different government and non-profit clients, offering services in more than 100 villages.
• Led the Delhi government's USAID funded Mission Convergence project providing consolidated government health benefits through partner non-profit organisations.

Jan 2005 – Jan 2008 · 3 yrs · Bangalore, India

• Part of the early engineering team at Bangalore office, joining initially as an intern and later as an employee:
• Worked on search quality to identify product review pages better, including various AI / ML classification and clustering problems
• Led Bangalore Engineering team that worked on dMarc (Google's acquired Radio Ads platform), including integrating it with the Google Ads platform, adapting it to work with India specific radio station consoles, designing and implementing conditional ad targeting, and setting up a Radio lab in Bangalore engineering office.
• Led Bangalore Engineering team of 6 engineers transitioning Youtube monetization, piloting initial invideo Ad prototypes.

Jan 2001 – Jan 2004 · 3 yrs · Cambridge, Massachusetts, United States

• Pursued Master’s in Computer science. Research assistant and core team member of Computer Architecture Group's Raw microprocessor group. The group designed, fabricated, built application and system software for path breaking multicore grid microprocessors. Took advanced courses in distributed systems, machine learning, algorithms.