Why Google ADK signals matter for AI companies
Google Agent Development Kit (ADK) is Google's open-source framework for building multi-agent AI systems on top of Gemini models and Google Cloud infrastructure. Developers building with ADK represent a high-intent segment: they are actively building production AI agents, evaluating agent orchestration frameworks, and making infrastructure decisions around Vertex AI, vector databases, and LLM observability tools. GitLeads monitors GitHub for ADK signals and delivers enriched lead profiles to your sales stack in real time.
Google ADK signals to monitor on GitHub
Stargazer signals — repos to track
- `google/adk-python` — official Google ADK Python package
- `google/adk-samples` — ADK sample applications and templates
- `google/generative-ai-python` — Gemini API SDK for Python
- `google-gemini/cookbook` — Gemini API examples and guides
- `google/vertex-ai-samples` — Vertex AI platform samples
- Related: `langchain-ai/langchain`, `crewAIInc/crewAI` — ADK developers often evaluate alternatives
Keyword signals in GitHub Issues, PRs, and code
- `google.adk.agents` — core ADK import in Python code
- `from google.adk` — ADK module import pattern
- `google-adk`, `google_adk` — package references in requirements.txt, pyproject.toml
- `SequentialAgent`, `LlmAgent`, `BaseAgent` — ADK agent class usage
- `google.adk.tools`, `FunctionTool`, `google_search_tool` — ADK tool patterns
- `AgentTool`, `session_service`, `MemoryService` — ADK session/memory API
- `vertexai.init`, `GOOGLE_CLOUD_PROJECT` — Vertex AI integration signals
Lead profile: what GitLeads delivers per ADK developer
{
"name": "Marcus Chen",
"email": "mchen@startupxyz.com",
"github_username": "marcuschen",
"company": "StartupXYZ",
"followers": 287,
"top_languages": ["Python", "TypeScript"],
"signal_type": "keyword",
"signal_context": "from google.adk.agents import SequentialAgent — PR: 'Add multi-step research agent'",
"repo": "startupxyz/research-agent",
"profile_url": "https://github.com/marcuschen"
}Who builds with Google ADK
ADK developers are a specific persona: Python-first AI engineers building production agent systems for enterprise or SaaS customers. They are evaluating:
- LLM observability and tracing tools (Langfuse, Helicone, Phoenix)
- Vector databases for agent memory (Weaviate, Qdrant, Chroma)
- Agent testing and evaluation frameworks (PromptFoo, DeepEval, RAGAS)
- Deployment infrastructure (Cloud Run, Vertex AI Endpoints, Modal, RunPod)
- Multi-agent orchestration platforms
- Enterprise AI compliance and guardrails tools
Routing ADK leads into your sales stack
- HubSpot: create contacts with `google_adk` persona tag and route to AI-focused AE sequence
- Slack: real-time alert for any ADK repo stargazer with 100+ followers
- Smartlead / Instantly: sequence referencing their ADK usage — "building multi-agent systems on ADK?"
- Clay: enrich with LinkedIn title; filter for Founding Engineer, AI Lead, Staff Engineer
- Salesforce: log under `AI Agent Builder` segment with Gemini/ADK custom fields
ADK vs. other agent frameworks — targeting the switcher
Monitor both ADK repos and competing frameworks to catch developers evaluating multiple options:
- Stars on `google/adk-python` AND `openai/swarm` → cross-framework evaluator, highest intent
- Issues comparing ADK to LangGraph → sophisticated buyer making architectural decisions
- Mentions of `adk` + `langchain` in the same PR → migration or integration builder