AI Engineer Associate (Agentic AI Workflows)
Build and maintain Agentic AI workflows, AI agents, CRM and lead-management automations, business-document generation, API integrations, evaluation systems, and human-in-the-loop safeguards.
About the role
Prewise Digital is looking for an AI Engineer Associate to build and ship Agentic AI-enabled workflows across our digital platforms. You will work directly with the Founder / Head of Product and Technology on live, business-critical systems, turning manual marketing, sales and service processes into reliable, measurable AI automations. Our platforms span Agentic AI-enabled workflows, CRM, lead management, marketing and business-services platforms. You will design, build, evaluate and maintain AI agents and automations that capture and qualify leads, update CRM records, support marketing execution, and generate business documents, with clear points where humans stay in control.
Roles & responsibilities
- Develop Agentic AI-enabled workflows for CRM, lead-management, marketing and business-services platforms (for example capture, qualify, score and store leads, then trigger follow-up or human handover).
- Build proposal and document automation: generate, review and refine proposals, agreements and business documents from structured inputs and approved templates.
- Design prompts, tool/function calling and structured (schema-validated) outputs for reliable agent behaviour.
- Define and implement evaluation and guardrails for AI outputs: labelled test sets, accuracy and failure analysis, prompt-injection resistance, human-in-the-loop checkpoints and audit logs.
- Integrate agents with CRMs, databases, email and messaging channels through REST APIs and webhooks.
- Implement validated workflow and state transitions in code, so that the AI never bypasses business rules.
- Monitor cost, latency and quality of LLM usage; compare models and prompts and recommend improvements.
- Document workflows, prompts and decisions; take part in code reviews and follow structured development practices.
- Collaborate with product, marketing and sales stakeholders to turn business needs into automation requirements.
Technology stack
Candidates are free to choose their tools. Experience with any one of the options in each row is acceptable.
| Category | Technology (any one of) |
|---|---|
| Languages and frameworks | Python; Next.js (TypeScript) |
| LLM APIs | OpenAI, Claude, Gemini, or open-source models |
| Agent frameworks / automation | LangGraph, n8n, or similar |
| Data and CRM | SQL, Supabase, or a basic CRM / database |
| Integrations | REST APIs and webhooks |
| Version control | Git and GitHub |
Required skills
- 2–4 years of relevant experience in AI/ML engineering, LLM application development, automation or software development.
- Hands-on experience building with LLM APIs, including prompt design, tool/function calling and structured outputs.
- Working knowledge of Python or TypeScript/Next.js.
- Experience with at least one agent or workflow framework or tool (LangGraph, n8n or similar).
- Familiarity with REST APIs, webhooks and SQL.
- Understanding of evaluating LLM outputs, reducing hallucinations and basic AI safety practices.
- Basic Git and GitHub workflow knowledge.
- Strong problem-solving mindset and clear written communication.
Good to have
- Experience with CRM or lead-management systems and marketing automation (email, social, ads).
- Experience with Supabase or another hosted Postgres database.
- Exposure to observability tools (for example Langfuse or LangSmith).
- Experience with document generation or RAG-based workflows.
- Experience handling multilingual or noisy user inputs.
Education
Demonstrable skills (projects, GitHub, deployed automations) matter more to us than degrees.
Selection process
Four stages, designed to assess real work.
Resume screening
We review your resume together with your GitHub and project links to understand what you have built.
Take-home assignment
A short, realistic Agentic AI workflow task that reflects the day-to-day work of the role.
40-minute walkthrough
You walk us through your assignment and prior work, and we discuss design, evaluation and guardrails.
Panel discussion
A conversation with the Founder / Head of Product and Technology and the wider team about fit and expectations.
What you will gain
- Hands-on ownership of Agentic AI workflows used in live business platforms.
- Direct exposure to product and AI leadership in a founder-led environment.
- Experience across CRM, lead-management, marketing and business-services automation.
- Practice in building evaluation, guardrails and responsible AI controls for real use cases.
- A fast learning environment with room to grow as the platforms scale.
How to apply
Apply through the form on this site so your resume and role-specific answers reach the hiring team together. Prefer email? Send your resume and links to careers@prewisedigital.com with the subject line “AI Engineer Associate - Tuni”.
Apply online