AI tools mapped across 17 stages of the software lifecycle
AIM Research, Oct 2026AIM RESEARCH · LANDSCAPE · NOVEMBER 2026
Mapping AI across the software development lifecycle 2026
Where agents now do the work, and where humans still lead. We placed 144 AI tools across 17 stages of software development, from the issue tracker to incident response, and marked the decisions that still sit with people.

THE SHIFT BEHIND THE MAP
of developers use a coding agent every week, and 68% use one daily
JetBrains Research, May–Jul 2026name a terminal agent as their main AI coding tool, up from 21.3% in January
First Page Sage, Oct 2026of developers use one of the four largest tools, up from 71.2% in January
First Page Sage, Oct 2026WHERE HUMANS STILL LEAD
Five decisions still sit with people
Agents now work in most stages of the lifecycle. These are the points on the map where a person still makes the call.
Decide what ships
Release timing and audience stay a human call, even where pipelines run themselves.
See it on the mapApprove merges
Agents review every pull request, and a person still signs off the merge.
See it on the mapHold incident command
AI SRE agents triage and investigate, while people lead the response.
See it on the mapSet priorities
Agents now pick up tickets, and humans decide which ones matter.
See it on the mapSet direction and guardrails
Engineers define what agents build and the limits they work within.
See it on the mapEXPLORE THE MAP
Every tool, by stage and by how much it does on its own
Filter by section or column, or search for a tool. Hover a tool for its stage, column and any change of ownership.
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
OBSERVABILITY10 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
INCIDENT RESPONSE10 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
CLOUD COST MANAGEMENT4 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
HOST13 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
FEATURE FLAGS6 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
CI/CD6 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
COMMIT & REVIEW CODE10 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
TEST & QA8 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
SECURITY9 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
DEBUG & GET HELP7 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
WRITE DOCS6 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
COPILOTS
AGENTS
WRITE & MAINTAIN CODE15 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
COPILOTS
Agents
Tools that take a task and finish it on their own
AGENTS
BUILD APPS FROM PROMPTS6 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
DEVELOPER PORTAL & SANDBOXES9 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
IDES & ENVIRONMENT9 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
TERMINAL9 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
ISSUE TRACKER7 tools
Incumbents
Established platforms, now shipping AI features of their own
Challengers
AI-native tools, with a person still steering
Agents
Tools that take a task and finish it on their own
THE DEALS BEHIND THE NAMES
Many of these tools now sit inside larger companies
Cursor
$60B all-stock deal
Ona
Formerly Gitpod, runs Codex agents
Windsurf
Now Devin Desktop
Statsig
$1.1B
Graphite
Code review and stacked PRs
Eppo
$220M
Base44
Prompt-to-app builder
HashiCorp
Maker of Terraform
Splunk
Part of Cisco
Split
Now Harness FME
Spot
From NetApp
CloudHealth
With VMware
Incumbents are established platforms that now ship AI features of their own. Challengers are AI-native tools built around a person who still steers. Agents take a task and finish it on their own, with a person reviewing the result.

RESEARCH & ADVISORY
Talk to AIM Research about your AI engineering stack
AIM Research evaluates AI, data and cloud vendors for enterprise buyers through the PeMa Quadrant, published since 2018, and helps teams shortlist the tools that fit.



