ByteAsk Secures $1 Million in Pre Seed Funding
ByteAsk, a startup dedicated to AI coding agents, has raised $1 million in a pre seed funding round led by Y Combinator and Entrepreneur First, with contributions from angel investors. The founders of ByteAsk previously established and sold the legal AI startup LawSutra AI to Manupatra, an online legal research platform.
Funding Purpose and Strategic Plans
The funds acquired will be utilised for product and infrastructure development, as well as hiring in engineering and research roles. The startup also aims to invest in GPU compute and training data. Additionally, ByteAsk plans to establish enterprise-grade security, privacy, and on-premises infrastructure for its customers.
Introduction to ByteAsk
Founded in the current year by Anirudha Kulkarni and Pratyush Saini, ByteAsk is focused on developing AI coding agents tailored for C and C++ coding environments. The platform specifically assists engineers engaged in mission-critical applications spanning defence, aerospace, robotics, high-frequency trading, finance, embedded systems, automotive, and semiconductors.
Addressing Challenges in C and C++ Development
ByteAsk aims to tackle the constraints of existing AI coding tools for C and C++ development, particularly in sectors where reliability, performance, and accuracy are paramount. Industry estimates indicate that C++ accounts for approximately $400 billion in annual engineering salary expenditure. The startup perceives agentic coding as a rapidly growing opportunity, valued at around $10 billion.
Targeting Large Enterprises
The initial focus of ByteAsk will be large enterprises in the high-frequency trading, automotive, and embedded systems sectors, with plans for future expansion into adjacent markets. The startup claims that engineers interact with its platform six times more intensively each day compared to a comparable open-source coding agent, and the number of weekly active users is reportedly doubling each week.
Internal Benchmark Performance
According to its internal benchmarks, ByteAsk’s grounding environment enabled a smaller model to effectively resolve 89% of firmware engineering tickets, surpassing a mere 61% resolution rate for the best frontier model tested without the grounding environment.
Innovations in C++ Training
Furthermore, ByteAsk is advancing a post-training approach for C++, with intentions to unveil a language model specifically optimised for C++ within the next six to eight months.