D.
Get in touch
Diyorjon Olimjonov

AI/ML Engineer LLM Applications & Voice AI

Diyorjon Olimjonov

Get in touch Work Resume
About

AI/ML Engineer with 2 years of experience. I worked on the data science team of a cybersecurity startup (raised $4M at seed stage), and for the last 6 months I have been building LLM and Voice AI applications.

As a side project, I recently built WENES CRM, an AI-powered CRM for the CIS / Uzbekistan market, now testing its MVP.

I am open to opportunities where I can work on LLM and Voice AI applications for production at scale, and do R&D to bring valuable solutions.

Experience

WENES-AI

Fergana, Uzbekistan

Founder & Founding Engineer

Jul 2026 - Present
  • Founding engineer on WENES CRM: took an open-source CRM core, built on it, and localized it for the Uzbekistan market, integrating AI to semi-automate the workflow from conversations to sales, helping SMEs grow efficiency and revenue.
  • Built a RAG pipeline for multilingual documents (including tables and code-switched text): hybrid retrieval (embeddings, SQL search, reranking) plus online / offline evaluation, improving Recall@1 to 0.82 and MRR@10 to 0.9 from baseline.
  • Implemented a self-learning system that enriches the knowledge base over time through FAQ, knowledge, and performance gap detection.
  • Built a temporal-memory knowledge graph (Graphiti) to track user context over time, plus an agent tool-use loop that reads from and writes to local databases and external CRM / ERP systems.

HumbleBeeAI

Incheon, South Korea · On-site · IT Services & IT Consulting

AI/ML Engineer: Data Science | LLM Applications | Voice AI

Nov 2024 - Jul 2026
  • kumush.ai: led and built a multi-tenant AI voice platform over a 2-month engagement: real-time LiveKit voice agents with telephony in Uzbek, Russian, and English plus a workflow / automation layer, owning planning, architecture, and latency optimization as project lead and main engineer.
  • wakil.ai: semantic-caching R&D for a legal AI platform (10,000 monthly active users): a semantic FAQ cache layer targeting about 10% of query traffic to cut LLM API costs without hurting answer quality.
  • Cybersecurity startup: engineered a production device-fingerprinting and identity-resolution system (Go, PostgreSQL) with TLS / network fingerprinting (JA4+) and multi-layered VM / emulator detection, shipped to 1M+ users; built adversarial-testing and behavioral-simulation research (human-like mouse / keystroke modeling) to harden bot detection.
  • GEO/SEO startup: built an n8n-based AI scoring agent using OpenAI embeddings and classification to categorize AI citation source types, and an analytics POC pipeline turning that into visibility metrics.

Emaar Hospitality Group & Rove Hotels

Dubai · On-site

Customer Service Agent

Dec 2020 - May 2022
  • Home-base call operator for all internal and external calls, then a year at Front Desk.
  • Handled up to 80 arrivals and departures per shift in peak seasons; maintained end-of-day and auditing reports on night shift.
  • Became an "up-seller" in Jan 2022; awarded "Extra Mile Q1" in 2021.
Stack

Voice AI & Conversational Systems

  • LiveKit
  • Deepgram
  • Speech-to-Text
  • Text-to-Speech
  • Telephony · SMS

GenAI & LLM Engineering

  • OpenAI API
  • LangGraph · LangChain
  • RAG
  • Graph & vector DBs
  • AI agents
  • MCP
  • Evals
  • Prompt engineering
  • Semantic caching

Automation

  • n8n
  • Make

Data Science & Analytics

  • Python
  • Pandas
  • NumPy
  • SQL
  • BigQuery
  • Time series · classification
  • Behavioral modeling

Programming & Infrastructure

  • Go
  • FastAPI
  • PostgreSQL
  • Docker
  • GCP
  • AWS
Selected work

01 / 07 · 2026

WENES CRM: AI-Powered CRM for the Uzbekistan Market

Took an open-source CRM core, built on it, and localized it for the Uzbekistan market, integrating AI to semi-automate the workflow from conversations to sales and help SMEs increase efficiency and revenue.

  • RAG
  • Evals
  • AI Agents
  • Agent Orchestration
  • Knowledge Graph
View project →

06 / 07 · 2024

Behavioral Simulation for Decentralized Detection Systems

Probed the detection layers of a decentralized, blockchain-based security subnet with realistic behavioral-simulation models: custom kinematic and statistical models that scored a perfect 1.0 across challenge versions.

  • Application Security
  • Behavioral Biometrics
  • Kinematics
  • Statistics
View project →

07 / 07 · 2025

AI Scoring Agent: GEO/SEO AI Visibility Analytics

An n8n-based AI scoring agent for a GEO/SEO startup, using OpenAI embeddings and classification to categorize the source types AI answers cite, plus an analytics POC pipeline that turns them into visibility metrics.

  • n8n
  • AI Agents
  • OpenAI Embeddings
  • Classification
  • Data Analysis
View project →
Background

Inha University

Incheon, South Korea · Sep 2022 - Jun 2026

B.B.A., International Business and Trade · GPA 4.1 / 4.5

Relevant coursework

International Finance and Banking · Logistics and Supply Chain Management

Cross-major coursework

Algorithms and Data Structures · Object-Oriented Programming · Computer Security · Discrete Mathematics · Introductory Engineering Mathematics

Certifications & training

  • IELTS 7.0
  • TOPIK Level 4
  • CyberOps Associate, Cisco Networking Academy
  • Mathematics for Machine Learning, Coursera
  • Introduction to Machine Learning, Coursera

Languages

  • Uzbek (native)
  • English (fluent)
  • Korean (conversational)
  • Russian (limited proficiency)
Contact

Hiring for LLM or Voice AI work? Let’s talk.

AI agent Or message my AI agent: it knows my work, stack, and availability.
diyorjon.me@gmail.com Telegram Resume