Senior AI/ML Engineer at Aubrant Digital

Jyväskylä, Central Finland, Finland -

Full Time

Start Date

Immediate

Expiry Date

05 Aug, 26

Salary

0.0

Posted On

07 May, 26

Experience

5 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Azure OpenAI, RAG, Python, Azure AI Foundry, Document Intelligence, LangChain, Agent Orchestration, Prompt Engineering, Azure AI Search, LLM Evaluation, Vector Databases, Semantic Kernel, LangGraph, HITL Design, SOC 2 Compliance, IaC

Industry

Software Development

Description

Job Description: As a Senior AI/ML Engineer at Aubrant Digital, you will build the AI capabilities that power platforms across our client engagements. You will own reusable AI skills shared across projects: Document Intelligence pipelines, RAG over enterprise document libraries, GPT-4o reasoning chains for summarization and analysis, classification and field extraction with GPT-4o-mini, and agent orchestration. You will partner with Technical Leads, data engineers, and backend engineers to build accurate, auditable, confidence-gated AI workflows for regulated workloads. What qualifications make you an Aubrant Senior AI/ML Engineer? An AI engineer who designs for accuracy, auditability, and human oversight rather than impressive demos. A clear communicator who can explain prompt design, RAG architecture, and accuracy trade-offs to engineers, architects, and client stakeholders. Comfortable operating with ambiguity, capable of building skills in domains where the right answer requires domain expertise to validate. A mentor who raises the bar through prompt review, evaluation design, and pattern guidance. Customer-obsessed and outcome-focused, treating accuracy thresholds, HITL design, and audit trail as features that protect regulated work. Responsibilities: AI Skills & Document Intelligence: Build reusable AI skills consumed across engagements: Document Intelligence, document summarization, data normalization, anomaly detection, matching engines, and compliance test runners. Design and train custom Document Intelligence neural models for client-specific document types. Implement RAG over enterprise document libraries using Azure AI Search with hybrid vector + keyword retrieval and semantic ranking. Reasoning, Agents & Evaluation: Build LLM reasoning chains using Azure OpenAI (GPT-4o for complex reasoning, GPT-4o-mini for high-volume classification) with prompt versioning and guardrails. Design agent orchestration in Azure AI Foundry for multi-step workflows: extract, search, reason, and generate output with tool-use grounding. Build evaluation harnesses, accuracy thresholds, and drift detection; tie outputs to confidence-gated HITL review tiers. Production AI, Compliance & Mentorship: Implement audit trail patterns for AI-assisted workloads: prompt/response logging, evidence chains, and SOC 2 aligned event sourcing. Operate AI Foundry deployments, manage PTU vs. token-based billing decisions, and monitor accuracy and cost in production. Mentor engineers on prompt engineering, RAG design, agentic patterns, and evaluation; contribute to Aubrant's AI engineering standards. Qualifications: Bachelor's Degree in Computer Science, Machine Learning, or a related discipline, or equivalent experience; MUST be proficient in written and spoken English (85%). 5 to 8 years of professional engineering experience with at least 3 years building production AI / ML systems. Expert-level proficiency in Azure AI services, including Azure OpenAI (GPT-4o, GPT-4o-mini, PTU and token-based billing), Azure AI Foundry, Document Intelligence (custom neural models), and AI Search. Expert-level proficiency in RAG and agent design, including hybrid retrieval, semantic ranking, prompt versioning, guardrails, evaluation harnesses, and confidence-aware HITL design. Strong proficiency in Python for AI/ML development, including modern frameworks for LLM applications (LangChain, LangGraph, Semantic Kernel, or equivalent). Hands-on experience with Document Intelligence custom models, including training, evaluation, and production deployment of neural extraction models. Experience designing AI workflows for regulated environments: audit trail, prompt/response logging, accuracy thresholds, and drift detection. Working knowledge of Medallion data architecture, vector databases, and embedding pipelines. Solid Git, code review, and engineering standards discipline; experience with trunk-based development and IaC for AI deployments. Experience in financial services, professional services, or other regulated industries is a plus. Experience with .NET interop or polyglot AI service ecosystems is a plus. Excellent analytical and problem-solving skills; strong communication, collaboration, customer orientation, innovation mindset, and adaptability under ambiguity.

Responsibilities

Build reusable AI capabilities including Document Intelligence pipelines, RAG architectures, and agent orchestration for regulated workloads. Mentor engineers on AI standards and implement audit trails and evaluation harnesses to ensure accuracy and compliance.