Build your team of AI agents.

Automate decisions with evide

We're a Senior AI Product Studio. We build evidence-first agents that connect to your content and systems to search, synthesize, and support decisions and automation.

Request Agent Discovery (1 hour) Know more

✓ Organize and collect knowledge
✓ Explain real time data and metrics
✓ Take intelligent actions with automation

Senior AI Product Studio

Vanlog's mission is to unlock the value of your knowledge and workflows with custom AI agents.

We help businesses, public organizations, and research teams automate processes, accelerate decision-making, and empower people by connecting AI directly to your content and systems.

From discovery to deployment, Vanlog guides you through every step: opportunity mapping, agent and retrieval design, integration, evaluation, and handover. We deliver solutions you own and control.

Our team brings deep experience in AI, product design, and software engineering. We work hands-on with your experts to ensure every agent is robust, transparent, and aligned with your goals.

Evidence-first by default: answers with citations, predictable behavior, and clear limits. If you want to see what Vanlog can do, just get in touch.

Request Agent Discovery (1 hour) See our use cases

Turn company knowledge into an actionable asset.

We ingest documents, PDFs, and shared folders (and, when needed, email threads and chat logs) into a unified, searchable knowledge base.
Teams get evidence-first answers with citations, plus summaries and checklists that follow your procedures.
Designed to stay up to date with your sources: no manual copy/paste knowledge management.

Automate workflows with AI agents.

We build AI agents that execute repeatable workflows by calling your APIs and tools—turning requests into actions.
Typical automations include ticket creation/updates, report generation, data entry, and notifications.
Evidence-first by default: agents log what they did and why, and can cite the sources they used.

Leadership & AI Innovation Support

Strategy support to identify high-impact AI agent opportunities
Workshops to map knowledge/processes and define an Agent MVP
Enablement to scale AI agents with evidence-first practices and open source

Services

Agent Discovery (1 hour)

A structured 60-minute session to turn an AI agent idea into a concrete plan: use case, sources, integrations, risks, and next steps.

Custom AI Agents

We build custom AI agents end-to-end: a fast MVP to validate the use case, then engineering to production with integrations, guardrails, logging, documentation & knowledge transfer.

Workflow Automation Agents

AI agents that execute repeatable workflows by calling your tools and APIs—turning requests into actions (tickets, reports, data entry, notifications).

Leadership & AI Innovation Support

Strategy support to identify high-impact agent opportunities, workshops to define an Agent MVP, and enablement to scale with evidence-first practices and open source.

Agent Ops & Improvement

Evaluation, monitoring, logging, and continuous iteration to keep agents reliable as sources, tools, and processes change.

Training

AI is the future. Do not just buy a service or a tool: learn, and bring the AI competencies into your company and your teams.

Some projects

Erum2020

We are main organizers of eRum2020, the bigger European R Users Meeting of 2020 - a four days conference & workshop event with +100 international contributors. eRum2020 is a totally virtual event with free access for everyone.

Temperature fluctuations

Together with professor J.K. Dagsvik and S.H. Moen, we have worked for years on a research project focused on understanding the fractal properties of the temperature fluctuations. This section contains data, results, paper and resources for temperature analysis

Multi DoE

Open source package for multi-objective design of experiments and optimization in R. It implements different strategies for multi-objective optimization. Multi-criteria design of experiments algorithm that simultaneously optimizes up to six different criteria.

Template by Bootstrapious. Ported to Hugo by DevCows.