AnetLab — in brief
What this is. We build what runs. Software development studio. Digital products and AI/LLM systems on .NET 8/9/10, integrations, legacy modernisation.
Who. Victor Buzin, founder and principal engineer. 16+ years in enterprise. Ex-Head of Software Development at SEFE (formerly Gazprom Germania), 12+ engineers.
What runs in production:
- Product: Contextum — the document layer between an agent and an LLM
- Open source: Delibera, Agent Shaker, Marp Play (github.com/techbuzzz)
- Enterprise (under NDA): Cloud Intranet, SecureDocs VDR, SecureDoc Storage
Stack. .NET · .NET Core · Semantic Kernel · LangChain · RAG · MCP · Vue 3 · React · Azure · AWS · PostgreSQL · Kafka
How they work. Free product interview → commercial proposal within a day → development in iterations → knowledge handed over to your team.
Contact. project@anetlab.com · github.com/techbuzzz
Blog posts. /blog — senior engineering practice on .NET, AI/LLM, RAG.
Contextum — documents to Markdown for an AI agent
A service that prepares documents for an LLM: converts them to Markdown, compresses the context and reports what was lost on the way. A loss registry, Go and .NET, Docker.
Delibera — collective decisions with LLMs
Several models discuss a question in rounds, critique each other’s answers, and a Chairman weighs the arguments into a final verdict. RAG, MCP, streaming output.
Agent Shaker — coordination for AI agents
Register agents, give them real tasks, coordinate them over MCP. Milestones, shared markdown context, live updates and A2A delegation.
Marp Play — Markdown into slides
An online viewer for Marp presentations. Paste Markdown, get slides. Open source, and it runs locally.
About the studio
We build what runs.
We build and ship digital products and AI/LLM systems on .NET
One senior engineer and a proven team. Sixteen years in enterprise development, the last five in AI and LLM.
We design, write the code, ship it to production and hand the knowledge to your team. The engineer who does the work is the one you talk to.
01 — Services
You can order this now
AI/LLM system, end to end
RAG · agents · quality evaluation
A system that works in production: RAG with evaluation, agents over MCP, orchestration and monitoring. A measurable result on your own data.
- Data audit and architecture choice
- RAG pipeline and an evaluation set
- Agents and orchestration
- Deployment, monitoring, cost control
Digital product, end to end
From idea to release
Research, design, development, launch. Backend on .NET or Node.js, frontend on Vue 3, React or Blazor.
- Research and strategy
- Interface design
- Backend and frontend
- Infrastructure and CI/CD
Integrations and automation
We connect your systems
SAP, Datareon, 1C, Microsoft 365, TESSA, DELO, SKB Kontur and others. We remove manual data transfer and the gaps between systems.
- Process and master-system discovery
- Data exchange design
- Implementation and migration
- Integration support
Technical audit
We find the bottlenecks
A review of code and architecture: performance, scalability, security, tech debt. A report with priorities.
- Code and architecture review
- Profiling
- Security audit
- Improvement plan
Legacy modernisation
We update an outdated stack
.NET Framework and WebForms onto a modern stack. No business process downtime, no data loss.
- Solution audit
- Migration plan
- Phased port
- Sign-off
- Team training
CTO-as-a-service
A senior engineer in your team
Technical leadership part-time: decisions, hiring, review, team growth. Without adding headcount.
- Architecture decisions
- Help with hiring developers
- Team mentoring
- Working with tech debt
02 — Portfolio
Client systems in production
Names and details under NDA. A full case study on request.
Corporate Communications Platform
An intranet on a microservices architecture with Microsoft 365 for an international energy company.
SecureDocs VDR
A Virtual Data Room for secure transfer of confidential corporate documents to governing bodies: RBAC, MFA, watermarking, geo sharding, AES encryption, a complete audit trail.
SecureDoc Storage
ECM with distributed storage and advanced encryption.
03 — Stack
Technologies we use
04 — Process
How the work runs
Interview
Free. We work through the task, the goal and the constraints. You get a preliminary estimate.
Architecture
We design the solution, fix the scope and the definition of done. The plan is visible before we start.
Iterations
Short cycles with a demo. Something working every one to two weeks.
Handover
Deployment, monitoring, documentation, and the knowledge transferred to your team.
05 — Contact
Tell me about the task
We work through the task and the goal. After the call you get recommendations on how to prepare the project and a preliminary estimate within a day.