n8n

LangGraph

Retell AI

Make.com

Voiceflow

Telnyx

Claude AI

Hermes

Available for new projects

ELISHA
ABRIHAM

I build AI-powered automation systems that save companies thousands of hours — from n8n voice SaaS platforms and Retell AI agents to LangGraph pipelines and production scrapers.

Get in Touch

100+

Automations Built

20+

ML Projects

30+

Clients Served

Scroll

Who I Am

About Me

Elisha Abriham

Backend & AI Automation Engineer based in Addis Ababa, Ethiopia. I build production-grade intelligent systems — from n8n SaaS platforms and Retell AI voice agents to LangGraph orchestration and ML pipelines.

I've automated entire business workflows for companies in Canada, the US, and Europe — turning complex AI research into reliable tools that deliver measurable ROI.

Years Experience

100+

Automations Built

20+

ML Projects

30+

Clients Served

Core Competencies

n8n & Make.com Automation93%

Python & FastAPI95%

LangGraph / LLM Engineering88%

Retell AI / Voice Agents87%

PostgreSQL & Data Engineering85%

Docker / DevOps82%

Currently focused on: Building n8n-powered SaaS backends, deploying Retell AI voice agents at scale, and orchestrating LangGraph multi-agent systems for enterprise automation.

Automation Arsenal

Tools I Wield Daily

Production-grade automation platforms I use to build real systems for real clients.

n8nMake.comRetell AIVoiceflowTelnyxLangGraphHermesClaude AIOpenAISupabaseStripeAirtableBubbleApifyApollo.ioPythonn8nMake.comRetell AIVoiceflowTelnyxLangGraphHermesClaude AIOpenAISupabaseStripeAirtableBubbleApifyApollo.ioPython

⚙️

Workflow Automation

n8n

Built 50+ production workflows — lead pipelines, CRM sync, voice agent auto-provisioning, billing automation, and cross-platform data flows.

🔗

iPaaS

Make.com

Visual automation connecting Stripe, Airtable, HubSpot, and custom APIs in complex multi-step business scenarios.

📞

Voice AI

Retell AI

Deployed AI voice receptionists for real businesses — auto-provisioned with custom LLM prompts, phone numbers, and Stripe subscriptions.

🎙️

Conversational AI

Voiceflow

Designed voice and chat agent flows for lead qualification, appointment booking, and customer service across multiple industries.

📡

Cloud Communications

Telnyx

Integrated programmable SMS, voice bridges, and call routing — including emergency escalation pipelines and retry systems.

🕸️

LLM Orchestration

LangGraph

Orchestrated multi-agent workflows with stateful graphs — CRM automation, code review pipelines, and RAG-powered decision systems.

🧬

LLM / Tool Use

Hermes

Used NousResearch Hermes models for structured function-calling, JSON extraction, and agentic tool use in local inference pipelines.

🤖

AI Reasoning

Claude

Integrated Anthropic Claude for complex reasoning, code generation, document analysis, and multi-turn agentic conversations.

What I've Built

Projects & Research

AI automation systems, ML research, backend APIs, and data pipelines — shipped to production.

📞

FeaturedAutomation

AngleAlways Answer — AI Voice SaaS

Full-stack SaaS platform that lets businesses deploy AI voice receptionists in minutes — no technical setup required. Built the complete automation backend handling agent provisioning, call routing, billing, and real-time sync across multiple third-party services. Serving multiple business clients across North America.

Retell AIn8nStripeBubbleTelnyxSaaS

View Live →

⚡

FeaturedAutomation

MagixWorx — Speed To Lead Platform

AI-powered lead engagement system that bridges incoming leads to sales reps within seconds. Webhook triggers an LLM-routing engine that simultaneously dials the lead and the closest available salesperson, plays personalized audio, and merges both parties into a live call. Includes Google Maps + Apollo lead scraper, KBO-based company data enricher, and a WhatsApp real-estate appointment booking agent with Google Calendar and Gmail integration.

n8nOpenAIApolloApifyVoiceflowGoogle Sheets

View Live →

🏗️

FeaturedAutomation

SuperGiant — Content Architecture Agent

AI agent platform for content teams that takes a client brief via webhook, fetches brand context from Supabase, and runs a multi-step LLM pipeline to produce structured content architectures and ready-to-publish copy. Built on n8n with OpenAI, deployed as a production SaaS at supergiant.ca.

n8nOpenAISupabaseAI AgentContent Generation

View Live →

✅

FeaturedAutomation

Qwalify — Client Portal Automation

End-to-end automation layer for the Qwalify client portal. Automates onboarding flows, candidate qualification pipelines, and status updates — eliminating manual hand-offs between recruiters and clients. Live at qwalify-client-portal.vercel.app.

n8nAutomationRecruitmentPortalWorkflow

View Live →

📈

FeaturedML

Time-Series Forecasting & Portfolio Optimization

Benchmarked ARIMA, SARIMA, and LSTM on Tesla (TSLA), BND, and S&P 500 (SPY) daily price data from Yahoo Finance. LSTM dominated — TSLA MAE 1.56 vs ARIMA's 2.34. Applied Markowitz Mean-Variance Optimization on the LSTM forecasts to derive an optimal allocation: TSLA 40%, BND 30%, SPY 30% that maximises return per unit of risk.

LSTMARIMASARIMAPythonyfinanceMarkowitzPyTorch

GitHub →

🔐

FeaturedML

Fraud Detection System — FinTech ML

Binary classification pipeline for real-time financial fraud detection. Trained on credit card transaction data enriched with IP-to-country geolocation features. Served via a Flask REST API (/predict, /health), tracked with MLflow, and containerised with Docker for production deployment.

scikit-learnMLflowFlaskDockerPythonClassification

GitHub →

Toolkit