Senior Full Stack Developer – Generative AI, Agents

Techrx Recruiting

. Design and evolve complex workflows built on the Google Agent Development Kit (ADK), including specialized skills/tools, agent sessions and state, and long-term memory through RAG .

Posted 9/29/2026full-timeSão Paulo • BrazilSeniorWebsite

Core Competencies

Role fit

Core Competencies

Use this summary to align your resume positioning with the role.

Demonstrates advanced proficiency in Python, including asynchronous programming and performance optimization, while effectively designing and evolving complex workflows using the Google Agent Development Kit. Expertise in implementing observability for AI applications and optimizing agent frameworks is essential.

Highest-signal resume keywords

Advanced Python ProficiencyExperience with Google ADKRESTful API DevelopmentAWS Systems ArchitectureObservability Instrumentation

ATS Keywords

Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills

Asynchronous ProgrammingPerformance OptimizationFastAPIUvicornClean CodeAutomated Testing with PytestInfrastructure as Code with TerraformDynamoDBOpenSearchLLM Observability

Soft Skills

Technical LeadershipCode ReviewCommunication

Tools & Technologies

Vertex AIAmazon BedrockGitHub ActionsDatadogECS/Fargate

Industry Keywords

Agent OrchestrationRAG StrategiesStreaming (SSE)Session PersistenceVector Memory

Tech Stack

Tools & technologies

AWSCloudDynamoDBJavaScriptNext.jsPythonTerraformTypeScript

About the role

Key responsibilities & impact

Design and evolve complex workflows built on the Google Agent Development Kit (ADK), including specialized skills/tools, agent sessions and state, and long-term memory through RAG
Define communication and state-management standards between components of agentic systems
Lead initiatives to improve latency and reduce inference costs using Gemini models through Vertex AI
Optimize prompt engineering, streaming (SSE), and caching strategies
Design data flows for agents with session persistence and expertise in DynamoDB, as well as vector memory in OpenSearch
Design and maintain sandboxing integrations for executing agent-generated code using Amazon Bedrock AgentCore Code Interpreter
Define technical KPIs for agent quality, dashboards, critical alerts, and evaluation pipelines in Datadog LLM Observability
Implement and evolve input/output guardrails, PII redaction, and access policies
Serve as a technical reference for Mid-Level and Junior Developers and conduct code reviews focused on security and clean architecture for AI

Requirements

What you’ll need

Advanced proficiency in Python, including asynchronous programming/asyncio, advanced typing, and performance optimization
Experience with modern dependency managers such as uv and Poetry
Advanced proficiency in developing and consuming RESTful APIs and streaming (SSE)
Required experience with FastAPI and Uvicorn
Strong knowledge of Clean Code, automated testing with pytest/pytest-asyncio, Git, and CI/CD pipelines with GitHub Actions
Experience with agent orchestration frameworks such as Google ADK, LangChain/LangGraph, or equivalent technologies
Ability to customize or extend agent frameworks
Experience consuming LLMs through cloud provider APIs, including embeddings and RAG strategies
Proven experience instrumenting observability for AI applications, including distributed tracing and LLM Observability
Experience with AWS systems architecture, including ECS/Fargate, API Gateway, Cognito, and S3
Experience with Infrastructure as Code using Terraform
Knowledge of non-relational databases such as DynamoDB and single-table design
Knowledge of vector search/storage solutions such as OpenSearch, pgvector, or similar technologies
Advanced English proficiency for conversation, reading, and writing
Nice to have: Previous experience with Google ADK, Datadog (ddtrace/LLM Obs), OpenTelemetry, Langfuse, Arize, AWS Bedrock services, TypeScript/Next.js, feature flags/OpenFeature, blue/green or canary deployments, LiteLLM, and generative UI protocols such as A2UI

Benefits

Comp & perks

CLT employment arrangement