Maximilian Asam

AI Engineer

max@asam.dev
Augsburg, Germany
asam.dev
github.com/AsamMax

I'm Max, an AI engineer from Augsburg. I build agentic LLM systems for enterprise and public-sector clients, most recently AIS.chat, an open-source assistant that thousands of teachers use every day.

Practical experience

September 2022 – present

XITASO GmbH — AI Engineer

AI Engineer | Jan 2025 – present · Software Engineer | Mar 2024 – Dec 2024 · Working student | Jan 2023 – Mar 2024 · Intern | Sep 2022 – Jan 2023. The AI Engineer role was introduced in January 2025, built around the LLM and RAG work I was already doing.

AIS.chat: national AI assistant for schools, commissioned by the German federal government and the federal states (open source). Six-figure registered teacher base, thousands of daily active users. Team of six engineers, two of them AI engineers.

  • AIS.chat — co-designed and built the agentic layer: function calling over teacher-uploaded documents (PostgreSQL + pgvector, permission-aware access), web search, and single-page retrieval, so answers are grounded in real sources.
  • Built a unified interface across four LLM providers, enabling per-task model selection.
  • Contributed to evaluating generation quality (LLM-as-a-judge with manual review), which informed model, prompt and retrieval choices.
  • Delivered features end to end, including the Next.js frontend, under strict data protection requirements for a public-sector system.
  • Multivac chatbot — RAG assistant making a large corporate corpus and wiki conversational. Team of four (two from XITASO, two from the customer).
  • Enforced per-user permissions at retrieval time, so users only get answers from content they are allowed to see.
  • Multivac HMI — step-by-step migration of an industrial frontend codebase to Vue.
  • Further client projects (under NDA).
October 2020 – August 2022

Delta 4 Software Solutions OHG — Software Engineer

Part-time, alongside studies.

  • Developed WordPress websites and an Android app.
August 2017 – July 2020

SEWOBE AG — Apprentice Software Developer (Fachinformatiker Anwendungsentwicklung, IHK)

  • Developed modular cloud software and customer websites in an agile team, working with relational databases.
  • Led a team of four apprentices in building a module for the company's core software.

Education

October 2020 – March 2024

Technical University of Applied Sciences Augsburg — B.Sc. Computer Science

  • Final grade: 1.8
  • Thesis (1.7): “Using a Conversational Agent for Information Retrieval from an Organisational Wiki”, in cooperation with XITASO. Thesis seminar: 1.3.
  • Student tutor for introductory programming courses.
August 2017 – July 2020

Berufsschule 7 Augsburg — Vocational school alongside apprenticeship at SEWOBE

  • IHK final exam (Fachinformatiker Anwendungsentwicklung): 94/100
2017

Fachoberschule Augsburg — Fachabitur (Technik)

AI / LLM

Retrieval-augmented generation
Agentic systems & function calling
LLM evaluation (LLM-as-a-judge)
Multi-provider LLM integration
Prompt engineering
Embedding models
Guardrails & safety models
Fine-tuning
LangChain

Engineering

TypeScript / JavaScript (React, Next.js, Angular, Vue, Drizzle)
Python (SQLAlchemy)
C# (Entity Framework)
Java
Kotlin
SQL (PostgreSQL + pgvector, MySQL, MariaDB)
REST API design
Automated testing
Linux
GitHub Actions CI/CD

Cloud & Infrastructure

Azure (Azure OpenAI, Entra ID, AI Search, Key Vault)
Docker
Kubernetes (Helm)
Observability (Grafana, Prometheus, Sentry)
GCP

Languages

German (native)
English (fluent)