Full-stack · Backend-leaning · Voice AI

Production systems where latency, scale, and correctness all matter at the same time.

Voice and real-time media pipelines, microservices that survive contact with the real world, and the AI plumbing that ties them together.

See selected work Get in touch
  • Backend & APIs
  • Voice AI & Telephony
  • Real-time media
  • Cloud & DevOps

Where there is depth

Expertise

A working toolkit, not a checklist. Each area has been shipped, debugged, and operated in production.

Backend & APIs

Long-running services and the contracts between them.

  • Node.js
  • Nest.js
  • Express.js
  • Python
  • FastAPI
  • REST
  • GraphQL
  • tRPC

Voice AI & Telephony

End-to-end voice stacks — from microphone to model and back.

  • STT / TTS
  • Web Speech API
  • OpenAI
  • SIP
  • Kamailio
  • RTPEngine
  • JsSIP
  • WebRTC

Real-time media

Moving audio and video through pipelines that have to keep up.

  • WebRTC
  • RTMP
  • RTSP
  • GStreamer
  • FFmpeg
  • Socket.io
  • Janus

AI integration

Wiring LLMs into real products — observably, safely, on someone else's behalf.

  • OpenAI SDK
  • Ollama
  • RAG
  • MCP
  • pgvector
  • Prompt design

Cloud & DevOps

From container to cluster, with traces all the way down.

  • AWS
  • Docker
  • Kubernetes
  • Terraform
  • ArgoCD
  • Jenkins
  • OTEL
  • CI/CD

Data & persistence

Picking the right shape of storage for the problem at hand.

  • PostgreSQL
  • MongoDB
  • Redis
  • DynamoDB
  • Prisma
  • TypeORM
  • Liquibase

Selected work

Sample engagements

A few representative systems. Anonymized — what matters is the shape of the problem, not the logo.

01

Telephony stack for voice AI

Multi-stage Docker deployment of a Kamailio SIP proxy and RTPEngine media relay feeding a voice-AI agent. Tuned host networking and WSS termination for low-latency speech roundtrips. RBAC across multi-tenant APIs.

  • Kamailio
  • RTPEngine
  • SIP
  • WebRTC
  • STT / TTS
  • RBAC

02

Self-hosted AI workspace

A central FastAPI backend serving two SPAs (a language-learning app and a markdown knowledge base), backed by Postgres with pgvector. A decoupled MCP server handles every LLM call — provider keys, request logs, OpenTelemetry traces and per-request event files all live behind one SSE boundary.

  • FastAPI
  • PostgreSQL
  • pgvector
  • MCP
  • OpenTelemetry
  • React

03

Healthcare microservices & CRM

CRM-facing services are extended, legacy data is migrated, code is reviewed (mostly backend, ~70–100% of the diff), and the architectural decisions shape the next quarter of work. Test coverage, CI/CD, Kubernetes via ArgoCD.

  • Nest.js
  • GraphQL
  • Medusa.js
  • MongoDB
  • PostgreSQL
  • ArgoCD

04

Video conferencing platform

Real-time video and audio using WebRTC, with backend services for voice recognition and translation. Infrastructure stood up with Terraform, CI/CD and Docker from a clean slate.

  • WebRTC
  • Janus
  • STT
  • Terraform
  • Firebase

05

SaaS back-office for media clients

Tools for enterprise media customers, integrated with HubSpot, payments and e-invoicing providers. Lambda pipelines did the heavy media processing; Jenkins and Docker ran the show.

  • Node.js
  • Express
  • Vue.js
  • AWS Lambda
  • Kafka
  • Jenkins

Contact

Let's talk

Backend, voice-AI and platform engagements. The fastest way to get in touch is email.

Email kefir266@gmail.com GitHub github.com/kefir266
Based in Warsaw, Poland