Company:** TaxDome
Employment type: FULL_TIME
About TaxDome
At TaxDome, we’re building the #1 practice management platform for accounting firms in the US and globally. Founded in 2017, we’ve grown into a 400+ fully remote team across 40+ countries, serving tens of thousands of businesses worldwide with millions of end clients.
How we work
You’ll be part of a globally distributed team built on trust, ownership, and self-management. We focus on outcomes over activity — prioritizing clear ownership, pragmatic decision-making, and accountability for results over rigid processes. Collaboration is central to how we operate: we communicate openly, involve the right people early, and continuously improve how we build products and teams together.
About this role
We're looking for a Senior AI Engineer to help us scale faster by designing and running production AI services - RAG-powered assistants, MCP servers, AI agents - and building smart automations for our teams. You'll choose the right tool for each problem, from Python services to low-code platforms like n8n, designing solutions that save time and boost efficiency.
This is a hands-on senior role: you'll own solutions end to end - from scoping a business problem and choosing the architecture, to building and integrating it, to shipping it securely and operating it. The role is as much about engineering judgment - deciding what to build, why, and with which trade-offs - as about building itself.
It's a fully remote role, we are hiring across the EU, with a preference for candidates working within CET (±3 hours).
What you’ll be responsible for
- Lead discussions on business challenges and propose solutions using LLMs, AI agents, custom services, or low-code tools - whichever fits best.
- Make and document architecture and tooling decisions: code vs low-code, build vs buy, model selection - with explicit cost, latency, security, and maintainability trade-offs.
- Design, build, and operate production Python services - including our retrieval-grounded assistant with RAG, database, auth, and CI.
- Build and tune RAG pipelines: embeddings, vector search, and retrieval quality evaluation.
- Design, build, and maintain MCP (Model Context Protocol) servers that connect internal tools, data sources, and APIs to LLMs and AI agents.
- Design, configure, and maintain AI agents for internal projects - including tool-calling, orchestration, and multi-step agentic workflows.
- Build and maintain automations in n8n.
- Design, test, and optimize prompts and multi-step prompt chains, including systematic evaluation (evals) of prompt and pipeline quality.
- Optimize cost and performance of AI workloads: caching, latency profiling, choosing the right model per task, token budgeting, basic load testing.
- Use AI coding agents as a primary implementation tool: decompose problems into well-specified tasks, review and validate generated code, and keep the bar on tests, security, and maintainability.
- Integrate external and internal services through REST APIs, webhooks, and OAuth flows.
- Build internal MVP tools on top of generative models to validate business hypotheses.
- Research, evaluate, and implement new automation and AI tools; run pilot projects with generative models.
- Maintain a structured prompt and workflow library, and document solutions for maintainability and handoff.
- Apply sound data-handling and security practices when working with internal systems, customer data, and credentials.
- Collaborate cross-functionally to understand needs and deliver impactful, reliable solutions.
Example project: building a feedback hub that collects and processes user feedback from multiple sources (CRM, surveys, forums, call recordings) into a single dashboard.
What you bring
Must-have
- 5+ years in IT overall, including 2+ years hands-on building applications, agents, and automations on top of LLMs - applying generative AI to real products and business tasks — and demonstrated seniority owning solutions end to end.
- Python - solid backend engineering: designing services, working with databases and APIs, writing tests, comfortable with Git, containers, and CI/CD. Not just scripting or extending existing automations.
- Hands-on experience with RAG and retrieval: embeddings, vector databases, tuning retrieval quality.
- Databases: SQL (Postgres) and key-value (Redis); familiarity with vector search (pgvector or similar).
- Experience building and operating AI agents and agentic workflows (tool-calling, orchestration); solid understanding of LLM principles, AI workflow design, and multi-step prompt chains.
- Demonstrated ability to justify technical decisions and reason about trade-offs (cost, latency, security, complexity) - not just implement what's asked.
- Experience with AI-assisted development workflows (Claude Code or similar); able to stay accountable for the quality of agent-generated code.
- Sound data-handling and security awareness (credentials, tokens, PII, access scopes).
- Self-driven, curious, and proactive in experimenting with new tools and technologies.
- English - Fluent: able to lead professional communication, handle correspondence, and participate fully in work meetings in English.
Nice-to-have
- Experience designing and building MCP servers (otherwise we expect a fast ramp on MCP).
- Experience with workflow automation platforms (n8n, Make, or similar).
- Experience integrating OpenAI, Anthropic, or other AI APIs.
- Experience integrating third-party services (Jira, Slack, Google, HubSpot, etc.) via REST APIs, webhooks, and OAuth flows.
- Experience with LLM observability and eval tooling (Langfuse, promptfoo, or similar).
- Experience running services on managed container PaaS platforms.
- Background in text/data processing.
- Experience with Google Workspace API, Slack API, CRM systems.
- Familiarity with data/security governance frameworks relevant to internal tooling.
What Success Looks Like
- First 90 days: ramp on our stack (n8n, Python services, Postgres/pgvector, Redis, our PaaS), ship one meaningful automation or AI tool into production, document it, and make at least one documented architecture decision (e.g., code vs low-code for a real task).
- First 6-12 months: own a portfolio of live automations, agents, and MCP integrations; establish reusable patterns and a well-maintained prompt/workflow library; measurably reduce manual effort across the teams you support and demonstrably optimize the cost/latency of at least one AI workload.
What we offer
- Competitive compensation, paid in USD, transparently shared before the first interview
- Fully remote work with flexible hours
- 30 paid days off annually, plus sick days as needed
- Health & well-being support
- Learning & development budget to support your professional growth
- English lessons reimbursement
- Co-working space reimbursement
- Company-provided equipment (conditions may vary depending on the role)
- A high level of autonomy and ownership in your work
- The opportunity to make a real impact in a fast-growing global SaaS company
- A collaborative, international team with a strong product mindset
Upon successful completion of the interview process and acceptance of the offer to join, an employment verification check will be conducted as part of pre-boarding — confirming job titles and dates of engagement with 2 of your previous employers. This is a required step for all new joiners.
