Armen Inants
Verified Expert in Engineering
Machine Learning Developer
Yerevan, Armenia
Toptal member since December 14, 2021
Armen is an artificial intelligence professional (PhD in AI from INRIA) with a diverse skill set ranging from machine learning and logic-based AI to research and publishing in top venues, and from developing innovative software. With 10+ years of experience in different areas of technology, Armen is primarily interested in designing knowledge-driven AI systems, document-centric workflow automation, retrieval-augmented generation (RAG), knowledge graphs, and ontologies.
Portfolio
Experience
- Artificial Intelligence (AI) - 10 years
- Ontologies - 9 years
- Software Architecture - 7 years
- Knowledge Graphs - 7 years
- RAG Architecture - 4 years
- ETL for AI - 4 years
- Agentic RAG Systems - 3 years
- Retrieval-augmented Generation (RAG) - 3 years
Preferred Environment
PyTorch, LangGraph, CrewAI, LangChain
The most amazing...
...project I've developed is a general auction handling system written in Haskell that allows organizing competitions for artificial smart bidders.
Work Experience
RAG Optimization Engineer
Hadi Debs
- Built an AI research platform that transforms scanned auction catalogues and articles into a knowledge graph, enabling complex multi-hop questions that traditional RAG cannot answer.
- Developed a next-generation retrieval system that combines semantic search, visual search, fuzzy text search, and relationship traversal to deliver grounded, traceable answers across text and images.
- Created a knowledge distillation pipeline that converts scanned PDFs into structured, provenance-backed data, resolves duplicates across sources, and applies consistent conflict resolution.
- Established LLM-as-judge evaluation frameworks for extraction and retrieval quality, measuring accuracy, completeness, citation quality, coverage, and relevance, with dashboards for batch testing and experiment comparison.
- Delivered internal tools for ontology design, knowledge exploration, and system inspection, accelerating development and making results easier to validate and use.
- Architected the platform for scalability and long-term evolution, supporting multiple customers, ontology versions, and continued product expansion.
Principal Symbolic AI Research Engineer
Morningstar
- Designed and developed a retrieval-augmented generation (RAG) system for predictive ESG analytics, enabling the company to reduce full-time equivalent (FTE) requirements.
- Led research projects, coordinating with stakeholders and technical teams to deliver insights.
- Oversaw the deployment of an AI-powered predictive analytics system on AWS, guiding the MLOps team throughout the process.
Expert Engineer
Inria
- Optimized a model checker of temporal properties of concurrent systems using static analysis techniques.
- Implemented the resulting algorithm that detects query containment within the Construction and Analysis of Distributed Processes (CADP) toolbox.
- Published a paper in the Artificial Intelligence Journal.
MATLAB Algorithmist
HeadSense
- Applied advanced signal processing techniques to extract features from clinical audio data and trained a neural network that predicts the patients' intracranial pressure (ICP).
- Contributed to the algorithm for the intracranial pressure (ICP) prediction in a biomedical startup backed by General Electric.
- Conducted sensitivity and specificity analysis to measure the performance of a novel ICP test.
Experience
SDK for Synthetic Document Generation
http://easydataset.aiGuideline-based Prediction
General Auction Handling System
MedBrain
I built a web service for querying the knowledge base.
Data Augmentation for Analytics
Zhop | Dutch Auction Platform
https://vimeo.com/241670818• Acted as the primary client contact, translating business needs into technical specifications, managing expectations, and providing regular progress updates.
• Designed, built, and deployed the end-to-end system, including the database schema, back-end services, real-time customer application, and admin panel for inventory and auction management.
TECH STACK
TypeScript, Node.js, PostgreSQL
Education
PhD in Mathematics and Computer Science
Université Grenoble Alpes (UGA) - Grenoble, France
Certifications
IBM RAG and Agentic AI
IBM
Build AI Agents using MCP
IBM
Agentic AI with LangGraph, CrewAI, AutoGen and BeeAI
IBM
Agentic AI with LangChain and LangGraph
IBM
Fundamentals of Building AI Agents
IBM
Build Multimodal Generative AI Applications
IBM
Advanced RAG with Vector Databases and Retrievers
IBM
Vector Databases for RAG: An Introduction
IBM
Build RAG Applications: Get Started
IBM
Develop Generative AI Applications: Get Started
IBM
Sequence Models
Coursera
Functional Programming in Haskell: Supercharge Your Coding
FutureLearn
Probabilistic Graphical Models 2: Inference
Coursera
Probabilistic Graphical Models 1: Representation
Coursera
Machine Learning
Coursera
Skills
Libraries/APIs
PyTorch, Vue, Node.js
Tools
MATLAB, GraphRAG, Hidden Markov Model, AWS Deployment
Languages
OWL, RDF, Python, JavaScript, SPARQL, Haskell, TypeScript
Storage
Graph Databases, Data Pipelines, PostgreSQL
Frameworks
LlamaIndex, LangGraph, Apache Spark, Streamlit, Agentic Frameworks, AutoGen
Platforms
Amazon Web Services (AWS), CrewAI
Paradigms
Functional Programming, Synthetic Data Generation, Management, DevOps, Model Context Protocol (MCP)
Other
Semantic Web, Complex Reasoning, Ontologies, RDFs, Research, Symbolic AI, Retrieval-augmented Generation (RAG), Prompt Engineering, Large Language Models (LLMs), Artificial Intelligence (AI), Software Architecture, RAG Pipelines, Agentic RAG Systems, RAG Architecture, Knowledge Graphs, ETL for AI, RAG Systems, AI Agents, APIs, Full-stack Development, Back-end, Front-end, Spatial Reasoning, Machine Learning, Architecture, Online Auctions, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), LangChain, Vector Databases, Neuro-symbolic AI, Data Engineering, LinkML, Supabase, FastAPI, Optical Character Recognition (OCR), System Design, Agentic AI, Agentic AI Systems, API Integration, Document Processing, Workflow Automation, Startups, Leadership, Algebra, Category Theory, Probabilistic Graphical Models, Markov Model, Markov Chain Monte Carlo (MCMC) Algorithms, Bayesian Inference & Modeling, Long Short-term Memory (LSTM), Recurrent Neural Networks (RNNs), Gated Recurrent Unit (GRU), Signal Processing, Linear Optimization, Networking, Gaming, Graph Neural Networks (GNNs), Science, Formal Methods, NLU, Deep Neural Networks (DNNs), Qdrant, Communication, Generative Artificial Intelligence (GenAI), ChromaDB, Information Retrieval, Multimodal GenAI
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring