📍 Bangalore // Toronto
I started my first company out of college, raised VC at age 22 and fell in love with building useful software products.
I look at life (and career) as a series of varied experiences of varying timeframes. No experience is irrelevant and no useful work is meaningless. Everything together compounds one's understanding of the world.
I'm a backpacker, currently at 23 countries. Vietnam is next on my list!
I started Lancify out of college with a close friend, Azan. An ed-tech company that taught freelancers and agencies SaaS skills like Shopify, Webflow, etc. We doubled as an extension SaaS companies' education & channel partner programs.
We built a team of 8, scaled to $250k ARR and partnered with top SaaS companies across the world. 30k people from 20+ countries took our free/paid courses. We had a blended completion rate of 60%+ and NPS of 91, which is almost unheard of in ed-tech. We've had students start from scratch and go on to build 6-figure agency businesses.
Lancify was backed by Blume, Techstars and 12 marquee angels from India and USA. We steered the company to a strategic exit in 2022.
I consult with large companies to launch new products, verticals or initiatives inside the org with the spirit of a startup. Like a fractional founding team on contract. I work with them closely on product, GTM and recruiting.
Polygon
Polygon and Flipkart teamed up to launch a web3 loyalty program application, Firedrops The fun part is that there is distribution to 200M Indians and all loyalty initiatives are to be rolled up into Firedrops.
I worked with leadership on product strategy, roadmap, ran the first few sprints and helped source tech partners + recruit 4 full-time hires.
Flipkart
Flipkart was acquired by Walmart at a valuation of $20B in 2018 and began investing heavily into innovation since. I worked with the "New Media Experiences" team to re-imagine how 3D/AR/VR could transform the mobile and web app experience.
I did product strategy, roadmap and put together an external partner team to execute an end-to-end 3D modelling → GenAI content creation → deployment pipeline.
A large pharma company
Generic pharma companies haven't been the best at adopting technology, but are keen to catch the AI wave in order to gain an edge in this hyper-competitive space with razor thin margins
I'm working with leadership to put together a cross-functional team to build AI chatbots that enhance efficiency of employees. We use RAG to deliver employees specialized assistants that help them get better at functions like New Product Selection, Out-of-Service Investigations, etc.
I will list relevant non-proprietary data in the form of vector embeddings on lore.market, my "Kaggle for embeddings" side-project.
A political party
I'm not very political but I certainly find the intersection of politics and technology intriguing. I also think there's tremendous opportunity to build in this space if you're in India, the world's biggest democracy + tons of highly skilled tech talent.
I'm working with state-level leadership to extract insights from ground-level intelligence, opinion polls & social media. We use RAG on a fine-tuned LLM that's able to contextualize tabular data with qualitative data (like micro and macro public opinion).
I will list relevant non-proprietary data in the form of vector embeddings on lore.market, my "Kaggle for embeddings" side-project.
/thesis
For individuals
People can access complex financial instruments and trading infrastructure permissionlessly. Anyone, anywhere can bypass the gatekeeping that we have in TradFi and devise unbounded trading & investing strategies.
For builders
DeFi projects can attack margins of centralized counterparts because they don't need to bootstrap essentials like liquidity, market maker relationships, lending power, etc. They can just build on top of existing protocols.
/work
Built a margin trading protocol on top of Aave & Uniswap. Trade with 3-4x leverage on longs and shorts, P&L dashboard and full transaction history. You can find the contract here and code here.
Check out the dApp at scrat.finance and blog post here.
1. farming perp funding rates
2. leveraged interest rate farming on stables
3. hedging IL with long straddle positions
Probably will clean up the code and open-source it.
the sxene (backed by NEAR) is a network of music artists, fans and enablers that aims to democratize artist monetization in the music industry. The token economics are very unique, it uses a bonding curve and a staking mechanism.
I contributed to the token economics, product strategy, wrote the wiki and helped close funding from NEAR.
/thesis
Short to medium term
AI does content (text, video & image) very well. Content is the best way for companies & individuals to build distribution. Massive opportunity to abstract away the mechanics of content creation & posting -- enabling humans to just focus on content strategy, personality and give AI original insights & ideas.
Long term
Cost of human capital will tend towards cost of energy because machines can do most work better than humans can. If we tweak the economic system right, this could mean that humans don't need to work, yet receive high income from the state. Utopia?
/work
1. Contextual, automated and very human-like bot that replies and likes posts according certain custom hashtags and voice & tone. Evades Instagram's bot detection when running from any residential IP.
2. Automated tool that takes an idea input, creates contextually relevant content with GPT-4, parses it through a Canva design template (reel, carousel or single post), schedules and posts to Instagram. (WIP)
1. Contextual, automated and very human-like bot that replies to tweets likes posts according certain custom hashtags, target users, keywords and voice & tone.
2. A team of AI agents that takes idea inputs as well as scrapes specified Instagram pages, creates contextually relevant & human-like tweet threads and single tweets with GPT-4, schedules and posts to Twitter. (WIP)
Currently testing on a few anon and D2C accounts, very pleased with the results! Github
Automated tool for creating, formatting, and publishing SEO articles from a Supabase DB. Uses GPT-4 for content generation and DALL-E 3 for image creation.
Working on automating populating DB directly from an input prompt. That process is currently manual but otherwise works really well -- publishes 1 article per minute.
A marketplace that hosts and lists datasets in the form of vector embeddings that can be accessed by any RAG app with 1-line of code. Starting out with just listing useful embedding knowledgebases from non-proprietary datasets I collect through my consulting.
Waitlist at lore.market