Artificial Intelligence in CaliforniaStartupsCodeiumAnnounced the completion of a $65 million round ofSeries B FinancingThe round was led by venture capital firm Kleiner Perkins, with participation from Greenoaks and General Catalyst.
Codeium said it will use the funds to grow its team across functional areas and expand its generative AI-powered coding toolkit, which has already authored more than 44% of new code commits for over 300,000 developers.
This support comes at a time when many enterprises and startups are seeking professional language models to accelerate their software development projects. Many well-known companies, including GitHub and Amazon, have entered this field, which is expected to be a $10.6 billion opportunity by 2030, growing at a compound annual growth rate of more than 25%.
Why is Codeium unique?
Software development has historically been a manual process, with developers faced with the challenge of writing code and being stuck in long development cycles.
However, when generative AI entered the scene, with the launch of ChatGPT and Codex, things moved much faster and technologists started working with smart assistants to automate and enhance every aspect of coding work from writing code snippets to enhancing actual project code and translating languages.
The move marks a major shift in software development, but even with AI, the generated code snippets are far from adequate. They are more like boilerplate examples of working solutions for specific use cases.
Additionally, most AI coding products do not integrate with the environment in which developers work, focus on only one model, and send code to the model over the internet, which can present security risks.
MIT graduates Varun Mohan and Douglas Chen experienced these challenges first-hand and decided to address them by launching Codeium, a security-focused LLM toolkit that provides intelligent code suggestions based on the code base. This provides developers with personalized builds to accelerate every task in their development workflow.
“We leverage proprietary AI models to integrate with your code baseup to dateIntegratedmaximum"Codeium supports over 70 languages and runs in over 40 integrated development environments including Visual Studio Code, JetBrains suite, Eclipse, and Jupyter Notebooks," Codeium wrote in a blog announcing its Series B funding.
The AI toolkit, which can be self-hosted or deployed as a SOC2 Type 2 compliant SaaS, integrates with existing source code management (SCM) systems with full repository context awareness. It includes in-editor features like auto-completion and chat, and has written more than 44% of new commits for customers, the company said in a press release.
However, developing LLMs and using them to support toolkits is notonlyCodeium also claims to have the hardware layer, the infrastructure for large-scale AI workloads that has been developed for years. This helps the company reduce the cost of delivering models to users to a fraction of that of other players in the industry.
“With this unique background, we’ve grown from zero to over 300,000 users in less than 15 months, while increasing the scale and quality of the proprietary Large Language Models (LLMs) that power our product, building an intuitive and industry-leading user experience, and personalizing recommendations for every developer task,” the company added in a blog post.
Plan to cover the entire software development life cycle
Over the past year, Codeium has provided its AI toolkit to “hundreds of companies,” including Atlassian, Anduril, Clearwater Analytics, and several Fortune 500 companies, to improve internal developer productivity. The company offers a basic version of its personal-centric toolkit for free, while a business-focused version starts at $12 per user per month.
In all cases, the company stressed that customers' code is neither stored nor used to train public systems.
Now, with Series B funding, the company plans to expand its engineering and sales teams and extend the platform to support a wide range of developer needs. Its goal is to go beyond writing and running code to address tasks at every stage of the software development lifecycle, including system design, code maintenance (especially legacy code migration), and solving complex problems such as security vulnerabilities.
“Today, we believe Codeium can make developers 20x more productive. While that’s impressive, we believe there’s more that can be done. We see Codeium as more than just a coding assistant, but a productivity multiplier that will make developers 20x more productive than just 20x,” the company noted.
Other notable companies working in this space include Replit, Github Copilot, Sourcegraph, Codiga, Amazon Sagemaker, and Google’s AlphaCode.