Today we are excited and proud to announce our seed investment in Portkey.ai and our partnership with Rohit and Ayush!
We are also excited to welcome a strong set of angel investors into the company, including Abhinav Asthana and Sanjeev Sisodiya of Postman, Adit Parekh of Freshworks, Ankit Gupta of Quizziz, Gaurav Manlocha of Stoa, Jake Seid of Ballistic Ventures, Manish Jindal of Cloudflare, Miten Sampat and Aakrit Vaish of Peercheque, Oliver Jay previously at Asana, Pranay Gupta of 91Springboard, Sandeep Krishnamurthy of AWS AI and Shyamal Anadkat of OpenAI. Many are from our network in the US and India.
Our investment thesis in Portkey is rooted in the team.
I met Rohit for the first time in 2012 in Pune, when he was working on his first SaaS startup Framebench. Rohit started Framebench soon after graduating from BITS Pilani in 2011. Framebench was acquired by Freshworks in 2016. At Freshworks, Rohit led product at Freshdesk (CX offering of Freshworks) and helped scale it over the next five years. Rohit and I kept in touch over the years, whether in Pune or Chennai or on zoom!
Rohit joined Lightspeed-backed Pepper* in 2021 to head product — Rohit was the initial adopter of GPT3 for Peppertype.ai (Content Creation SaaS) and helped scale the product to 500k users in the 2021-22 timeframe before the GPT-led boom really got going. Ayush has worked as head of engineering in various fast growing startups such as Pepper and PubNinja.
We’ve seen Rohit in action over the past decade and really admire his product thinking and fast execution capabilities, combined with commercial instincts. In Ayush, he’s found the perfect partner to build out the product.
Like it often happens, Rohit and Ayush encountered the problem they got passionate about at the previous company they were working at. They then decided to come out and start a company to solve this problem.
At Lightspeed, we have time and again seen second time founders building on the learnings of their previous experiences and on the back of new tailwinds. We have seen this second-time founder dynamic at play in Acceldata*, Supabase*, Zluri*, Rubrik*, Thoughtspot*, and dozens of other infrastructure software startups.
The tailwind is strong here – in the vast landscape of computing, new building blocks are a rare occurrence. Throughout history, the core concepts of networking, storage, and compute have remained steady. However, in recent years, a groundbreaking innovation has emerged – large AI models.
Enter Stable Diffusion*, Mistral.ai* and OpenAI’s ChatGPT, these large models represent a paradigm shift. By seamlessly integrating them into applications and software, developers can unlock functionalities previously unimaginable. From generating visual and text content to classifying existing data and establishing semantic connections, large models present an entirely new frontier in software development.
The catch, however, is that utilizing these powerful models is not a walk in the park. Most developers are not well-versed in machine learning engineering, and the scarcity of machine learning engineers only adds to the challenge. For millions of software engineers worldwide, hurdles abound – from figuring out model hosting to handling unexpected errors and evolving models over time. The absence of clean abstractions and user-friendly tools for large models adds another layer of complexity. The journey from proof of concept to production-readiness can be frustratingly slow and complex.
This is where Portkey comes in. Portkey.ai enables companies to seamlessly develop, deploy, and manage LLM-based applications, transforming a once complex and time-consuming process from POC to a quick and robust deployment. Portkey empowers engineering, product, and ML teams to focus on innovation, while providing essential observability, experimentation, and compliance support. More at www.portkey.ai.
As the usage of AI spreads from technology companies to the broader economy, we think this is a problem that every organization will need to solve. Onwards and upwards!
* Represent Lightspeed-backed companies