← Cases
United StatesTechnology and telecom2024 – ongoing

PubMatic

Generative AI coding tools for engineers

PubMatic reported a 35% rise in engineering productivity in the first half of 2024 with generative AI tools

WorkingCompany's own claimOfficial ↗

Problem

PubMatic wanted its engineers to ship software faster and tested whether generative AI tools could help with coding, unit tests and code review.

Approach

Engineers used GitHub Copilot for code generation and unit tests, Bito.ai as a code review agent, and ChatGPT for small scripts. The company said engineers accepted 20-30% of Copilot's generated code. One Private Marketplaces project was sped up by 80% through faster test case writing.

Results

35%rise in engineering productivity in the first half of 2024, per PubMatic's President of EngineeringCompany's own claim[1]
more than 80,000lines of code written with generative AI tools, per PubMaticCompany's own claim[1]
80%speed-up of one Private Marketplaces project through faster test case writing, per PubMaticCompany's own claim[1]

Takeaway

The productivity figure is PubMatic's own and the post does not say how it was measured.

Categories

Business function
IT and R&D
Innovation type
Process innovation
AI technology
Generative AI
Evidence
Company's own claim

Sources

Source summaries

[1]PubMatic30 Jul 2024

A blog post by PubMatic's President of Engineering, Mukul Kumar, on generative AI tools used by its engineers. It reports a 35% productivity rise in the first half of 2024, more than 80,000 lines of AI-written code and an 80% speed-up on one project.

Full details are in the original sources.

Added 2 Nov 2026 · Updated 2 Nov 2026
Open case page ↗