Skip to content
Available for select engagements

Production software and AI systems, engineered end-to-end.

Lead full-stack engineer and architect, based in Berlin. One of the first engineers at Parloa, pre-seed to ~$3B, building production AI years before the hype. I ship production software and verified, auditable AI agents for funded startups and the teams building them.

web & mobile appsAI integrationagentic systemsRAGcontext engineeringconversational AIpayments & billingcode & architecture auditstechnical due diligence
Alican Torun
Built and shipped at
ParloaAntlerCirplusVisuo
9+
Years building products
~$3B
Parloa, joined pre-seed
6 figures
Revenue, year one independent
Antler '23
VC-backed founder
Client stories

Real projects, real results.

Alican helped us architect our SaaS foundation and delivered hands-on full-stack work across every layer of the platform. He operates with the mindset of a technical co-founder, not a contractor.
Robert Werner
Robert WernerTechnical Co-Founder & CTO, Leapter
Alican built standalone data connectors for Zendesk and Slack, a CLI tool, and set up our entire repository architecture. His expertise in modern web technologies moved our product forward fast.
Onur Eken
Onur EkenCo-founder & CTO, Needle
Alican built our medical platform across three phases: document management, AI-powered consultation transcription, and a real-time voice and chat system. He gave us a foundation that can handle future growth.
Dr. Nils Zänker
Dr. Nils ZänkerCo-founder, Pocket Doctor
Alican's decade of experience blends deep full-stack and AI expertise with an entrepreneurial edge. He built entire products in days, significantly speeding up development.
Philipp Diller
Philipp DillerCTO, Kalasar

Due to the nature of startup work and NDAs, not all projects can be shared publicly. Reach out for more references.

How I work

Three ways I work.

Build

Production software, AI integration, and technical rescue. Full-stack development from architecture to deployment.

See how I build

Advise

Senior technical strategy on architecture, AI, and scaling. For founders who need an engineer's judgment, not just hands.

See advisory work

Empower

Lead offering

Agentic AI built into your product. Verified, auditable agents with real tool access and a full audit trail, not chatbot demos.

See AI systems
Agentic AI

An agent proposes. A verifier and a human decide.

Most AI demos answer confidently, even when wrong. I build systems that know when they are unsure. The agent reads and writes the language around a decision made on cited, versioned facts. A verifier blocks anything ungrounded, and high-stakes calls route to a human. Every run is a replayable audit trail.

answer_inquiry · replayable traceappend-only log
  1. 01INTAKEParse the request into structured parameters. Never guess. Ask when unclear.
  2. 02RETRIEVEPull cited, versioned facts from an inspectable claim store, not raw model memory.
  3. 03REASONCall real tools with system access. Assemble a structured, cited proposal.
  4. 04VERIFYA verifier checks every citation, blocks anything ungrounded, respects hard constraints.
  5. 05GATEAnswer, redirect, or escalate to a human when the stakes or uncertainty are high.

Cited, versioned knowledge

Facts live in an inspectable claim store, never baked into model weights. No fine-tuning.

Verifier hard-blocks

Hard constraints are walls, not penalties. The system is never confidently, silently wrong.

Replayable audit log

Every retrieve, tool call, verdict, and human decision is appended and reviewable.

Philosophy

Architecture-first agentic coding.

I start with the system map. When the model sees your architecture, it stops guessing and starts reasoning. The setup determines the output.

Structure before code

Architecture docs, diagrams, and dependency maps go into context first. The model reasons about where a change belongs instead of guessing.

Boundaries prevent coupling

When component boundaries are explicit, the model will not silently create dependencies across them. Structure is a guardrail.

The codebase is the prompt

How code is organized determines how well AI works with it. Clean structure produces clean output. Messy structure produces spaghetti on autopilot.

Learn more about my approach
Working with me

What you get.

  • Direct access. You talk to me, not a PM or a junior.
  • Predictable scope, predictable pricing.
  • Senior expertise across the full product lifecycle.
  • AI-augmented delivery, end-to-end.
  • Strategic input on architecture and product decisions, not just execution.
Engagements

How we can work together.

Project-based

Defined scope, clear timeline, fixed deliverables. Most engagements run 4 to 8 weeks.

  • Product and 0-to-1 builds
  • AI integration projects
  • Architectural rebuilds

Monthly retainer

Ongoing engineering partnership. Strategic input, execution, and AI integration as needed.

  • Architecture and product decisions
  • Code reviews and technical guidance
  • Hands-on execution, flexible scope

Advisory

Senior strategic input without full execution.

  • Architecture reviews
  • AI integration planning
  • Hiring and team scaling
How it works

From first call to shipped.

A free intro call to start, fixed scope before any work begins, and 30 days of support after handoff.

01

Intro call

A free 30-minute call to understand scope, constraints, and whether we are a fit.

02

Proposal and scope

Clear scope, timeline, and milestones agreed upfront. No hourly surprises.

03

Build

Weekly updates, async communication, and continuous delivery you can follow.

04

Handoff

Production deployment, documentation, repo transfer, and 30 days of support.

Available for select engagements

Working with funded startups and senior teams across Berlin, the US, and the EU.

Get in touch to discuss your project. I will respond within 24 hours.

Alican TorunBerlin, Germany