Meet the engineers accountable for your system.
Syntalith is an engineering team. The four engineers responsible for delivery and external technical review bring more than 22 years of software engineering experience combined. The company was registered in October 2025, after its engineers had already built and operated production systems for years. Your process scan is led by the engineer responsible for the architecture, integrations, or agent.
01Artem Lisovtsov
Co-founder · Applied AI/ML Architect
- Python
- Golang
- AWS
- AI/ML
- Voice AI
- AI agents
More than 6 years in production systems: backend, AI/ML, FinTech, and MediaTech
Co-founder of Syntalith; designs the architecture of every deployment and owns the AI/ML layer. He has more than six years of experience spanning custom CMS and CRM systems, e-commerce, high-traffic backends, and AI/ML. At Andersen Lab, he progressed from Senior Backend Engineer to AI/ML Engineer and AI Systems Architect. His work covered the full path of production AI systems: chatbots, model training and fine-tuning, private deployments, RAG platforms, workflow orchestration, and agents that use governed sources and tools. He built semantic search on AWS Bedrock, OpenSearch, and pgvector for production load, monitoring, and maintainability. He also developed FinTech and MediaTech systems serving hundreds of thousands of users a day, including for a forex broker and a European TV broadcaster. He runs Syntalith's own conversational systems and co-builds odbierze.ai, a specialist AI voicebot implementation agency.
- At Syntalith
- architecture of every deployment and the AI/ML layer
- Experience
- more than 6 years across full-stack, backend, high-traffic systems, and AI/ML
- AI/ML and GenAI
- chatbots; model training and fine-tuning; private deployments; RAG and semantic search; workflow orchestration; AI agents using governed sources and tools
- Systems and agency
- Syntalith conversational systems running in production and odbierze.ai, an AI voicebot implementation agency
- Certifications
- AWS Certified Machine Learning – Specialty; Claude Certified Architect, Foundations (Anthropic)
- Training
- five individual AI training sessions, including Kuba Koziej
- Named reference
- Kuba Koziej, CEO and co-founder of MoreGrowth and board member at Natu.Care: one-to-one training, a personal AI operator and team AI enablement.
02Serhii Ivanchatenko
Co-founder · Voice AI & Agent Systems Engineer
- .NET
- Voice AI
- AI agents
- LLM evals
- MSSQL/Oracle
- Qdrant
More than 7 years of commercial .NET work across PLM, CRM, and internal systems
Co-founder of Syntalith; together with Artem he runs the company and odbierze.ai, a specialist AI voicebot implementation agency. He has more than seven years of commercial .NET experience. At Trace One, he develops PLM modules and integrations, distributed processing, and CI/CD workflows in Azure. Previously, he worked on PLM systems for global food and consumer-goods manufacturers, with high-volume data-exchange APIs, MS SQL and Oracle, and asynchronous processing. He also built an internal CRM and engineering-automation platform for more than 100 users from scratch. At Syntalith, he builds voice AI and agent systems: telephony, CRM and calendar integrations, and the backend through which agents use client tools. He defines evaluation sets, tests scenarios and edge cases, reviews execution traces, and iteratively refines prompts, tools, and exception handling.
- Commercial experience
- more than 7 years in .NET, since 2018
- Domains
- PLM for global food and CPG manufacturers, CRM, enterprise
- At Syntalith
- voice AI and agent systems: telephony, integrations, MCP components, workflows, and Qdrant
- Voice AI
- lead engineer of odbierze.ai: call handling, calendar and CRM integrations, and conversation monitoring
- Evaluation and refinement
- case sets, scenario and edge-case testing, trace analysis, and iterative prompt and tool refinement
Trusted external auditors
A second pair of eyes on complex audits.
They run production LLM systems in large organizations. We studied computer science together at Lublin University of Technology. For complex audits and architecture decisions, they provide an independent review under their own names.
He has more than five years of IT experience, including more than two years of hands-on AI and LLM work. He leads a GenAI team at PwC Poland and was previously a lead software engineer at Asseco Business Solutions. He designs and deploys RAG and multi-agent systems, runs open LLMs on-prem, and leads fine-tuning and evaluation. He builds Python microservices on Kubernetes and automates infrastructure across AWS and Azure with Terraform and Helm.
- Experience
- more than 5 years across testing, software engineering, solution architecture, and technical leadership
- GenAI systems
- RAG from conventional implementations to multi-agent architectures, LangChain, LlamaIndex, and vector databases
- Models and evaluation
- on-prem hosting of open LLMs, fine-tuning, and evaluation of production outcomes
- Infrastructure
- Python, Kubernetes, Terraform, Helm, CI/CD, AWS, and Azure
- Python
- RAG / multi-agent
- LLMOps
- Kubernetes
- Terraform/Helm
- AWS/Azure
Spent four years at Asseco Business Solutions building and operating the backends of high-traffic systems. He designed APIs with FastAPI and Pydantic, asynchronous pipelines with Celery and RabbitMQ, and a RAG chat platform with Qdrant. He integrated LangChain, LangGraph, and MCP, owned Prometheus monitoring, and automated local vLLM deployments on Kubernetes with resource-aware scheduling and error handling.
- Production backend
- Python, FastAPI, Pydantic, PostgreSQL, and API architecture for high-traffic systems
- AI systems
- RAG, Qdrant, LangChain, LangGraph, and MCP with OpenAI, Anthropic, and local models
- Processing and monitoring
- Celery, RabbitMQ, Prometheus, unit testing, and integration testing
- Local models
- automated vLLM deployment on Kubernetes with resource-aware scheduling and error handling
- Python
- FastAPI
- RAG/Qdrant
- Kubernetes
- Celery/RabbitMQ
- vLLM
Business development
Who books the calls.
Daria usually handles the first message, qualification and scheduling. An engineer leads the technical conversation and process scan.
Daria Sandach
Business Development
LinkedInAt Syntalith since August 2025, she has run prospecting, qualification, and scheduling, including for odbierze.ai. She talks with owners about where a team loses time and context and whether an AI agent suits a given process; her background is business psychology. If you book a scan, she will usually coordinate it.
Meet the engineer who will lead your build.
- A 30-minute call with the engineer who would lead the work.
- A review of the processes that cost you the most time and money.
- A written summary: a possible direction, missing information and the next step.
Times are shown in your own time zone. We work with clients across time zones.
The scan chooses one process to assess, and within 2 business days you receive a recommendation, including when a simpler route is the better fit.
We start building after you sign the proposal.