БразилияSenior Deployed AI Engineer – Anthropic
3 нед. назад
БразилияSeniorOnsite
aidata engineeringdeploymentconversational interfaces
Senior AI Engineer specialized in Anthropic ecosystem, building AI products from idea to production including conversational applications and multi-step agents.
О компании
- is a next-generation data and AI consulting firm dedicated to accelerating the adoption of data and AI to create measurable business impact across the full enterprise value chain. We sit at the intersection of consulting, data science, AI technologies, data engineering, and digital transformation. We do not just advise — we build, implement, and deliver results our clients can measure. Our teams bring together consultants, data scientists, data engineers, AI engineers, analysts, and digital experts to solve complex business challenges with pragmatic, production-ready solutions. As continues to grow globally, we are building a team of entrepreneurial data and AI talent who can help clients move beyond experimentation and into scalable, governed, value-generating AI adoption.
Обязанности
- is looking for a Senior Deployed AI Engineer specialized in the Anthropic (Claude) ecosystem: an engineer who works embedded with our clients and takes AI products from idea to production. You will design and build the interfaces, services, and agentic systems at the heart of our client work, such as conversational applications over enterprise data, multi-step agents that automate business workflows, and the retrieval and data pipelines that support them. You will own your components end to end: the front end, the service behind it, the data and retrieval pipelines feeding it, the deployment, and the evaluations proving it works. This role combines deep, certified expertise in the Anthropic ecosystem (Claude models, the Claude Platform, and its agentic tooling) with the ability to deliver end to end. Beyond your platform specialization, you will be expected to work confidently across the full delivery lifecycle — full-stack development, data engineering, cloud infrastructure, evaluation, and client communication. You will work closely with our clients, with direct exposure from the start, and you will support the professional development of the engineers around you.
- Develop , schedule, and monitor data pipelines using Apache Airflow.
- Collaborate with data analysts, scientists, and engineering teams to ensure reliable data delivery.
- Optimize ETL processes for performance, reliability, and scalability.
- Maintain and improve data architecture, ensuring data integrity and security.
- Troubleshoot and resolve pipeline failures and performance bottlenecks.
- Provide technical leadership and guidance in the implementation of technical solutions, ensuring alignment with industry standards and compliance requirements.
- Document architectural decisions , strategies, and implementation guidelines for future reference.
- Lead by example , promoting engineering excellence and a culture of continuous improvement.
- Coach and upskill team members by providing technical guidance, pair programming, design reviews, and knowledge sharing.
- Build Full-Stack AI Applications, End to End You will build AI products across the entire stack, from interface to infrastructure.
- Develop user-facing interfaces in TypeScript/React and the backend services and APIs behind them in Python or Node.
- Implement agentic behavior: orchestration, tool and function calling, memory, and guardrails.
- Build retrieval-augmented generation (RAG) pipelines: ingestion, chunking, embeddings, vector and hybrid search.
- Connect AI systems to enterprise data and applications via APIs, semantic layers, and protocols such as MCP.
- Go Deep on the Anthropic Platform You will be the team's reference for the Anthropic platform.
- Design and build agentic systems with the Claude API and Claude Agent SDK: tool use, extended thinking, prompt caching, structured outputs, and citations.
- Build and deploy cloud-hosted agents with Claude Managed Agents: sandboxed code execution, memory, checkpointing, webhooks, multi-agent orchestration, and end-to-end tracing.
- Build and integrate Model Context Protocol (MCP) servers and clients to connect Claude to enterprise systems and data.
- Make Claude Code a core part of delivery — skills, subagents, hooks — and help clients adopt agentic development workflows.
- Deploy Claude in enterprise environments via the Claude API, Google Cloud Vertex AI, or Amazon Bedrock, selecting the right model (Opus, Sonnet, Haiku) for each cost, latency, and quality trade-off.
- Track Anthropic's releases closely and translate new capabilities into client value quickly.
- Make AI Systems Production-Grade Our standard is production quality: systems that are reliable, monitored, and maintainable.
- Write evaluation suites and regression tests for LLM-powered features, and monitor cost, latency, and quality in production.
- Apply solid engineering practice: version control, code review, automated testing, CI/CD, and observability.
- Deploy on cloud infrastructure (GCP, Azure, or AWS) using containers, serverless, and infrastructure-as-code.
- Build and maintain the data pipelines that feed AI systems, across warehouses, lakehouses, and vector stores.
- Work AI-Natively and Client-Facing Our engineers work AI-natively and represent directly with clients.
- Use agentic coding tools (Claude Code, Gemini CLI, Codex, Cursor) daily, with good judgment about verification and review.
- Communicate progress, trade-offs, and blockers clearly to clients and project leads.
- Support pre-sales when needed: scope solutions, build demos, and estimate effort with our partnership and consulting teams.
- Mentor junior engineers and contribute to internal accelerators, reusable components, and engineering standards.
- What We're Looking For Required Experience
- 3–5 years of experience in software engineering or data engineering, with extensive hands-on use of AI tools and LLM-based development over the past year (professional projects, internal initiatives, or substantial personal builds).
- Professional English proficiency (C1/C2 minimum) — mandatory. You will work daily with international clients and colleagues.
- Strong hands-on experience with the Anthropic ecosystem: Claude API, Claude Agent SDK or Claude Code, and MCP — ideally with experience taking at least one solution to production.
- Strong programming skills in Python and TypeScript/JavaScript, and experience building and consuming APIs.
- Experience with front-end development (React or similar) and at least one backend framework.
- Hands-on experience with RAG, embeddings, and vector search, and with at least one agentic framework (Claude Agent SDK, LangGraph/LangChain).
- Working experience with at least one cloud platform (GCP, AWS, or Azure); running Claude via Vertex AI or Amazon Bedrock is a plus.
- Fluency with agentic coding tools such as Claude Code, Gemini CLI, Codex, or Cursor.
- Experience building and maintaining data pipelines.
- Bachelor's or Master's degree in computer science, engineering, or a related field, or equivalent practical experience.
- Certifications An Anthropic certification is a strong differentiator at application. If you do not hold one yet, obtaining it within your first 2 months in the role is a requirement — sponsors the exam and gives you time to prepare
- Claude Certified Developer – Foundations (Anthropic Certification Program), or a higher-level Claude Certified Architect credential.
- Anthropic Academy completion certificates (Claude API, MCP, Claude Code) are valued as supporting evidence.
Будет плюсом
- Experience with MCP servers, multi-agent patterns, or LLM evaluation tooling (LangSmith, Langfuse, promptfoo).
- Experience with Terraform or CI/CD pipelines.
- Contributions to the MCP ecosystem, published MCP servers, or open-source agentic tooling.
- Key Capabilities A strong candidate will bring:
- Deep expertise in the Anthropic platform, combined with breadth across the full stack
- Owns features end to end, from interface to infrastructure
- Cares about evaluation and reliability, not just the happy path
- Communicates clearly with clients in demos, documents, and code review
- Client-facing mindset: understands client needs and translates business requirements into technical solutions
- Learns new tools and models fast, and shares what works
Условия
- Meal (VR)
- Free Office (work from home or anywhere you want!)
- Gympass
- Insurance: Life, Health, and Dental
- Bi-monthly Meetings (our “Get Together” where we meet to be together, with workshops, lectures, training, and a happy hour!)
- Woba (you can book coworking spaces anywhere you want!)
- Semi-annual evaluations (with opportunities for promotion)
Другое
- is the place to be: come and build the future of marketing
- Progress: every day offers new challenges and new opportunities to learn
- Culture: join the best team you could ever imagine
- Entrepreneurship: you will be joining a team of driven entrepreneurs. We won’t give up until we make a huge dent in this industry!