Jellyfish, a software engineering intelligence platform provider, today expanded its integrations to support additional AI coding tools, as engineering team juggle multiple AI coding assistants in their development toolsets.
The Jellyfish platform enables leaders to align engineering work with strategic business objectives. By analyzing engineering signals and contextual business data, Jellyfish provides complete visibility into engineering organizations, their work, and how they operate, the company said.
The Boston-based company announced Tuesday that it has added support for Cursor, Google Gemini Code Assist, and Sourcegraph to its AI Impact platform, joining its existing GitHub Copilot integration. The move reflects a growing trend among software organizations that are no longer betting on a single AI coding tool.
“Half of our customers are using multiple AI tools just in their IDE already,” Krishna Kannan, who leads Jellyfish’s product organization, told The New Stack. “That was really our cue that we needed to diversify beyond Copilot and see what else is being used.”
Founded in 2017, Jellyfish operates in what it calls the “engineering management” category, helping mid-market and enterprise software companies track productivity and efficiency across their development teams. The company serves organizations ranging from 25 engineers to more than 10,000, Kannan said.
Measuring Multi-Tool Impact
The new integrations enable engineering leaders to understand which AI tools are actually delivering value and where. While individual AI vendors provide usage data through their APIs, Jellyfish enriches that information with business context from project management systems like Jira, source control data from Git, and organizational charts, Kannan said.
“A lot of folks can understand adoption of AI impact through using the vendor tools,” he explained. “But if you want to understand productivity, you need to enrich it with your Git and Jira data.”
Based on measurements across more than 300 companies and 20,000 engineers, Jellyfish showed that AI coding tools are delivering significant productivity gains, including:
- 25% faster coding time
- 12% increase in pull request throughput
- 17% increase in roadmap work versus maintenance tasks
However, results vary by company and type of work, with infrastructure development showing less improvement than other coding tasks, Kannan said.
The Multitool Future
Jellyfish’s expansion reflects a broader shift in how organizations approach AI coding assistance. Rather than standardizing on a single tool, many companies are allowing — or even encouraging — developers to use different tools for different purposes, Kannan said.
“Our assumption is that the future will be multitool, and that different tools will have different specialties in different areas,” Kannan said. “It won’t just be always use Cursor, always use Copilot, but different types of needs will bring different results.”
This approach creates new management challenges. Engineering leaders need to understand not just whether their teams are adopting AI tools, but which tools work best for specific tasks and team members.
Hootsuite, a social media management platform, exemplifies this challenge. “As we rolled out Cursor across our engineering org, we were limited by their usage data and couldn’t effectively track adoption or identify productivity gains,” said Ciaran McAuliffe, SVP of software development at Hootsuite, in a statement.
With Jellyfish’s Cursor integration, the company now has “granular adoption data and the ability to benchmark against non-Cursor users,” McAuliffe added. “With clear signals on where productivity gains are happening, we can take learnings from the higher-performing teams and share them across the org.”
Growing Competition
Jellyfish faces competition from other software engineering intelligence vendors, including Swarmia, LinearB, and GetDx, though Kannan noted that homegrown solutions remain their biggest competitive challenge. Many organizations attempt to build their own measurement systems by tapping directly into vendor APIs.
However, the company differentiates itself by focusing on practical engineering leadership problems and providing actionable insights rather than simply displaying metrics. “We really root ourselves in what does the data show us? How can I make it actionable and bring insights based on that data?” Kannan said.
As AI coding tools continue to proliferate and evolve, companies like Jellyfish are positioning themselves as essential infrastructure for managing the transition. The challenge is no longer whether to adopt AI coding assistance, but how to do it effectively across diverse teams and use cases.
2024 Report
Meanwhile, in their research for the 2024 State of Engineering Management Report, Jellyfish found that 9 out of 10 engineering organizations are using GenAI. However, there’s a disconnect between leadership and their teams: 76% of executives believe their team has embraced AI, while only 52% of engineers say the same.
While some teams may survey engineers to learn how they use AI, sentiment alone is not enough. The key to effectively incorporating AI into your processes and securing the needed investment is an objective measurement of its actual impact, the company said.
The right metrics ensure that engineers are given impactful tasks and see the value in their work. Companies must be able to effectively measure their team’s adoption of AI and — more importantly — the technology’s impact on building better products and shipping them faster.
“We are all attempting to understand the true impact of GenAI coding tools. Leadership and the board are trying to understand how much value AI delivers. Our teams want to know how it will impact their career,” Kannan said in a statement released last year. “That’s why we built our Copilot Dashboard that accurately measures the impact of GenAI tools.”
The post Jellyfish Tracks AI Impact Across Four Major Coding Tools appeared first on The New Stack.
Software intelligence platform Jellyfish expands beyond GitHub Copilot to track Cursor, Google Gemini Code Assist, and Sourcegraph as engineering teams increasingly adopt multiple AI coding tools simultaneously.