sigmacode Assistant
A multilingual sales and knowledge assistant that only knows this website's content. Answers stream in your language.
Try it liveAI & Automation
AI that does real work in your business.
Assistants that answer from your own knowledge, pipelines that turn PDFs into structured data, and agents that automate multi-step work. Built for reliability, cost control and privacy.
Problems we solve
“Our team answers the same questions every day.”
An assistant grounded in your documentation that answers with sources — and says when it doesn't know.
“We type data from PDFs into our systems by hand.”
Document pipelines that extract validated, structured data with a review step where it matters.
“We tried a chatbot and it made things up.”
Grounding, citations, structured outputs and evaluation sets that make quality measurable.
“We're worried about cost and data privacy.”
Prompt caching, the right model per step, rate limits and GDPR-conscious data flows.
What we build
01
Assistants on your data
Multilingual chat assistants that answer from your documents and knowledge base (RAG), cite their sources and say when they don't know.
02
Document processing
Invoices, contracts and forms → validated, structured data for your ERP, accounting or CRM, with a human review step where it matters.
03
AI agents & workflows
Multi-step automations with n8n, Make or custom Python/Node — email triage, lead handling, report generation and more.
04
LLM integration
Claude, OpenAI or open-source models integrated into your existing product, with streaming UX, structured outputs and fallbacks.
05
Guardrails & evaluation
Rate limits, input validation, prompt-injection hygiene and evaluation sets so quality is measured, not guessed.
06
Cost control & privacy
Prompt caching, the right model for each step, and data handling designed for GDPR from the start.
What you get
- Working AI feature integrated into your stack
- Prompt and schema design, versioned in your repository
- Evaluation set and quality baseline
- Rate limiting, logging and cost monitoring
- Documentation for your team
- Full IP transfer
Try our AI demos live
These showcase demos run on Claude via the Anthropic API. Usage is rate-limited and your inputs are not stored.
Document Q&A with citations
Upload a PDF or paste text and ask questions. Every answer links back to the exact passages it's based on.
Try it liveInvoice data extractor
Turn an invoice PDF or photo into validated, structured data — table, JSON and CSV export.
Try it liveSmart Contract AI Reviewer
Paste Solidity or pick one of our showcase contracts and get a structured, severity-ranked security pre-review.
Try it livePackages
Transparent starting prices. Every project begins with a free scoping call and a written proposal.
Starter
AI Starter
A chatbot on your documents or one automation workflow.
from$2,500
- Assistant on your knowledge base
- Or one automated workflow
- Multilingual
- Rate limiting & logging
Growth
AI Growth
AI agent or multi-step automation with integrations.
from$6,000
- Multi-step agent or pipeline
- Integrations (CRM, ERP, email)
- Structured outputs
- Evaluation set
Enterprise
AI Platform
Multiple AI features across systems, with governance and evaluation.
Custom quote
- Several use cases
- Self-hosted model options
- Evaluation & monitoring
- Team training
Fixed scope, fixed price
AI Proof-of-Value Sprint
Find out in two weeks whether an AI use case is worth building.
Fixed price
$7,500
What's included
- Use-case scoping with measurable success criteria
- RAG or agent prototype on your own data
- Evaluation set with real questions and expected answers
- Cost model per request and per month at your expected volume
- Data handling under our AI data principles and a DPA
You receive
A go/no-go report with evaluation results, risks, the cost model and — if it's a go — a plan and estimate for production.
Free checklist
Scoping a RAG or AI assistant project
Use case, data sources, access control, retrieval, evaluation and costs — the questions to answer before building an AI assistant.
Your data in AI projects
Provider and region per project, EU options, no training on your data, PII redaction and a DPA.
How we handle your dataNot sure which package fits?
Answer five short questions and get an indicative budget range and duration for your project.
Selected work
Client work and showcase projects
Client platforms shown with permission, plus our own clearly labelled showcase projects you can try.
GAGA Menjačnica
A fast Next.js platform for a currency exchange and gold & silver dealer in Novi Sad.
Read case study
RCIID
A platform for an Austrian blockchain forensics firm that presents crypto transaction tracing to lawyers, authorities and victims.
Read case study
Sisak Green
A bilingual (DE/EN) investor platform for an industrial waste-to-energy project in Croatia.
Read case studyShowcase projects by sigmacode
- sigmacode Assistant
- Document Q&A with citations
- Invoice data extractor
- Smart Contract AI Reviewer
- Milestone Escrow
- Token + Vesting + Staking
- SigmaNFT Collection
- Gold-Backed Token (RWA)
- SigmaGovernor DAO
- Subscription Manager
Tech stack
AI & Automation
Cloud & DevOps
How a project runs
- 01
Discovery
A free scoping call to understand your goals, users and constraints. We ask the uncomfortable questions early.
- 02
Plan
A written proposal with architecture, scope, milestones, risks and a fixed price or milestone budget.
- 03
Milestones
We build in short iterations with working demos. You get repository access and can give feedback at every step.
- 04
Handover
Deployment, documentation and a walkthrough for your team. Full IP transfer on final payment.
- 05
Care
Optional retainer for maintenance, hosting, security updates and continuous improvement.
Frequently asked questions
Which AI models do you use?
We pick per use case: Anthropic Claude, OpenAI models or open-source models you can host yourself. Our live demos run on Claude.
Will our data be used to train models?
We use commercial API offerings whose terms exclude training on your API data by default, and we design data flows to minimise what leaves your systems.
What if the AI gets something wrong?
We design for it: grounded answers with citations, structured outputs validated against schemas, confidence checks and human review steps where errors are costly.
Can you automate a workflow across our existing tools?
Yes — with n8n, Make or custom code, connecting email, CRM, ERP, spreadsheets and internal APIs.
How long does an AI project take?
A focused assistant or one automated workflow can be production-ready in a few weeks. Multi-step agents and integrations across several systems take longer; the proposal includes a milestone plan.
What does it cost to run?
Running costs depend on usage and the models chosen. We estimate them during scoping and design for cost from the start, e.g. with prompt caching and smaller models for simple steps.
Can the AI run on our own infrastructure?
Yes. Where data must not leave your infrastructure, we can use open-source models hosted in your environment.
Let's scope your project
A short call is enough to tell whether and how we can help, and what it will roughly cost.
Prefer to write first? Write to us