Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Full-stack TypeScript developer shipping production AI systems on top of solid product engineering. I work end-to-end across React/Next.js frontends and Node/NestJS services, and go deeper than most on the operational side of LLMs: orchestration, retrieval, validation, and crash recovery. Recently I’ve built and led agentic workflows and orchestration systems (batch + real-time) with exactly-once, crash-idempotent processing, resilient job execution, and verification/validation pipelines. I also deliver multi-tenant web + mobile platforms and LLM-powered booking/conversational agents hardened with durable async processing, testing, and release gates.…Full-stack TypeScript developer shipping production AI systems on top of solid product engineering. I work end-to-end across React/Next.js frontends and Node/NestJS services, and go deeper than most on the operational side of LLMs: orchestration, retrieval, validation, and crash recovery. Recently I’ve built and led agentic workflows and orchestration systems (batch + real-time) with exactly-once, crash-idempotent processing, resilient job execution, and verification/validation pipelines. I also deliver multi-tenant web + mobile platforms and LLM-powered booking/conversational agents hardened with durable async processing, testing, and release gates.

Aggelo Sko(n)stantil(er)is

Full Stack Developer, AI Engineer, Back-End Developer, +3





Full-stack TypeScript developer shipping production AI systems on top of solid product engineering. I work end-to-end across React/Next.js frontends and Node/NestJS services, and go deeper than most on the operational side of LLMs: orchestration, retrieval, validation, and crash recovery. Recently I’ve built and led agentic workflows and orchestration systems (batch + real-time) with exactly-once, crash-idempotent processing, resilient job execution, and verification/validation pipelines. I also deliver multi-tenant web + mobile platforms and LLM-powered booking/conversational agents hardened with durable async processing, testing, and release gates.…Full-stack TypeScript developer shipping production AI systems on top of solid product engineering. I work end-to-end across React/Next.js frontends and Node/NestJS services, and go deeper than most on the operational side of LLMs: orchestration, retrieval, validation, and crash recovery. Recently I’ve built and led agentic workflows and orchestration systems (batch + real-time) with exactly-once, crash-idempotent processing, resilient job execution, and verification/validation pipelines. I also deliver multi-tenant web + mobile platforms and LLM-powered booking/conversational agents hardened with durable async processing, testing, and release gates.

Available to hire

Full-stack TypeScript developer shipping production AI systems on top of solid product engineering. I work end-to-end across React/Next.js frontends and Node/NestJS services, and go deeper than most on the operational side of LLMs: orchestration, retrieval, validation, and crash recovery.

Recently I’ve built and led agentic workflows and orchestration systems (batch + real-time) with exactly-once, crash-idempotent processing, resilient job execution, and verification/validation pipelines. I also deliver multi-tenant web + mobile platforms and LLM-powered booking/conversational agents hardened with durable async processing, testing, and release gates.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Beginner

Language

Greek, Modern

Fluent

English

Advanced

French

Intermediate

Work Experience

Founder & Full-Stack TypeScript Developer (Part-time) at Independent projects (HoloComm and Apollo CRM products)

May 1, 2025 - Present

HoloComm — Multichannel AI Booking Agent: case study architecture of a multi-tenant, omni-channel AI booking backend on NestJS 11 (domain-driven & hexagonal). Implemented integrations across OpenAI, Redis, MongoDB, Stripe, and Google Calendar, and messaging across WhatsApp/Telegram via Zernio’s unified API. Built a LangGraph conversational agent with tool/function calling, Zod-validated I/O, three-tier input guardrails, and regex-first bilingual routing to reduce LLM calls. Hardened with durable async processing (scheduled jobs, outbox delivery, race-safe atomic booking) and extensive case/fixture testing and evaluation with release gates. Apollo CRM — Pet Boarding & Hospitality CRM: built and maintained a production pet boarding CRM (Next.js App Router + MongoDB + TypeScript) used in a real business. Designed the financial layer tracking real transactions across payments (including reversals) and expenses; engineered room-based scheduling with real-time conflict/occupancy dete

Developer at Apollo on CRM (Pet Boarding & Hospitality CRM)

May 1, 2025 - Present

Built and maintained a production pet-boarding CRM (Next.js App Router + Server Actions, MongoDB, TypeScript) with ~109K LOC, 444 components, 284 server actions, and 32 models. Served a live business with 784 clients, 844 bookings, and 73 boarding rooms. Engineered room-based scheduling with real-time conflict/occupancy detection, drag-and-drop re-assignment, and Synchfusion + Google Calendar sync managing 2,356 scheduled appointment sync points. Designed financial data layer tracking ~€794K in real transactions across ~1,378 payments (including reversals) and ~€509K expenses over ~18 months of live operation.

Founder & Full-Stack TypeScript Developer (part-time) at HoloCom( / Holocomm)

May 1, 2025 - Present

Built a multi-tenant, omnichannel AI booking agent backend on NestJS 11 using domain-driven & hexagonal architecture. Implemented a conversational LangGraph agent with ~21 tool-calling functions, Zod-validated I/O, three-tier input guardrails, and regex-first bilingual routing that handles ~70% of turns with zero LLM calls. Hardened with durable async processing (28 scheduled jobs, outbox delivery, race-safe atomic booking) and extensive test coverage/evaluation at scale with release gates. Integrated with OpenAI, Redis, MongoDB, Stripe, Google Calendar, and messaging across WhatsApp, Telegram, and more via Zenrio unified API.

Senior Full-Stack Developer & AI Lead at LGMAR (LG Mar’s maritime email and operations product organization context)

November 1, 2024 - Present

Developed and maintained LGMAR’s maritime email and operations product across two surfaces: a legacy Omni web platform (~3,000 high-value users) and iTelex Plus, a cross-platform React Native app (4.5★, 1,000+ installs). Led AI feature work across both. Built an AI-agent bug-resolution pipeline with custom Claude/OpenAI agents across analyze→plan→execute→verify, parallelized via git work trees. Fixed 43 tracked bugs to completion with strong automation gates (high plan-validation landing rate; meaningful share of major/critical). Drove leading planning for an upcoming migration to React 18 + TypeScript.

Senior Full-Stack Developer & AI Lead at LGMAR

November 1, 2024 - Present

Developed and maintained LGMAR’s maritime email and operations product across two surfaces: the Omni web platform (legacy DHTML LX + JavaScript) used by ~3,000 high-value users and iTelyx Plus, its cross-platform React Native app (4.5★, 1,000+ installs). Led AI feature work across both. Built an AI agent bug-resolution pipeline (analyze → plan → execute → verify) using custom Claude agents in parallel via git work trees to plan-validate automations. Drove production fixes for many tracked bugs with high landing rates and automated plan-validation gates. Led planning for OMNI 2, including the upcoming migration to React 18 + TypeScript.

Senior Full-Stack Developer at Motion Hellas

August 1, 2024 - October 31, 2024

Led development of a production multi-tenant waste-management platform used by three organizations (AFIS Cyprus, ReBattery, Fotokyklos) tracking ~20,000 collection points. Built React + TypeScript single-page apps (SPAs) with advanced filtering, inline editing, and Excel export. Designed GraphQL APIs and role-based access control (RBAC) aligned with EU waste-tracking compliance requirements.

LLM Batch & Real-time Orchestration Engineer (Lead) — sole engineer; case study designer/builder at Forge

August 1, 2024 - October 31, 2024

Designed and built an entire provider-agnostic LLM orchestration system: a two-daemon orchestrator with nine worker loops supporting both LLM batch and real-time jobs across OpenAI and Anthropic. Implemented exactly-once, crash-idempotent processing that prevents lost or duplicated job runs across crashes/N-replicas. Sustained ~40 jobs/sec per replica (~3.4M/day) and scaled horizontally using PostgreSQL alone (no queue service), running large jobs via provider Batch APIs at roughly half standard token cost.

Education

BSc, Computer Science (GP A 7.0/10) at University of Piraeus

September 1, 2012 - February 1, 2024

BSc, Computer Science (Software Engineering) at University of Piraeus

September 1, 2012 - February 1, 2024

Qualifications

Add your qualifications or awards here.

Industry Experience

Software & Internet, Computers & Electronics, Transportation & Logistics, Travel & Hospitality, Media & Entertainment, Financial Services, Professional Services