Are you an A+ Series startup? Get first talent FREE OF CHARGE! Check if you qualify →
Hire as Freelancer
37 €
/hour
Not available
or
Hire as Employee
5875 €
/month
$
€
TALENT-67165
Mels
Machine Learning Engineer
Armenia
Seniority
Senior
Language skills
English B2
Hire employees directly using our Employer Of Record & Payroll tool:
- Recruitment fee, talent's one month salary
- EOR €199/month
Skills
Industry
Professional Summary
Candidate with over 6 years of experience in machine learning and software engineering, focusing on classical machine learning, reinforcement learning, and self-driving car research. Expertise includes training large-scale LLAMA models with reinforcement learning, handling models up to 70 billion parameters, and developing a document analysis product for financial analysts. Specialized in LLMs and Retrieval-Augmented Generation (RAG), including the development of a RAG system for code to enhance retrieval accuracy. Proven track record in leading teams, scaling machine learning efforts, and deploying models using AWS, Docker, and Kubernetes. Additionally, led the creation of a state-of-the-art educational application, achieving significant improvements in performance and user engagement. Skilled in a wide array of machine learning frameworks and technologies, including LangChain, PyTorch, AWS, Azure, MongoDB, and PostgreSQL.
Video of Talent
Portfolio
Education
Yerevan State University / Master's degree, Mathematics of financial systems
2017 - 2019
Yerevan State University / Bachelor's degree, Economics
2013 - 2017
Certifications and Trainings
Experience
Machine Learning Engineer
Industry: IT Services and IT Consulting
Sept 2024 - Present
Working on LLM/RAG evaluation and guardrails.
Machine Learning engineer (Contract)
Industry: AI
Feb 2024 - Jun 2024
Worked on a document AI application with the objective of creating a system to analyze unstructured handwritten documents and organize them into a predefined structure.
Lead Machine Learning Engineer
Industry: Software Development
Jul 2023 - Jan 2024
Helped the company achieve one of its two most important milestones following a Series A funding by fine-tuning Llama 270B and 13B models with Reinforcement Learning from Human Feedback (RLHF) and deploying them on AWS using Docker and Kubernetes.
Led a team of six engineers (ML, front-end, and back-end) to build a RAG-based "chat with PDF" application tailored for the financial industry, leveraging the fine-tuned model. The application gained active user adoption.
Founder, CEO
Industry: Software Development
May 2022 - Jul 2023
Developed a car highway lane-keeping system with 25x higher performance compared to 95% of current production vehicles while utilizing resources equivalent to legacy methods.
Engineered the system using an end-to-end deep learning approach.
Lead Research Engineer
Industry: Automotive and AI
Oct 2021 - May 2022
Directed a team of research, software, and hardware engineers in the development of a reinforcement learning-based self-driving car using human feedback.
Created a simulation agent capable of learning basic actions from eight examples and 40 minutes of training on a 4GB VRAM GPU.
Collaborated with the hardware team to develop a camera-only system for collecting real-world driving data, integrating the car's control sensors, including steering wheel angle, acceleration, and deceleration.
Lead Machine Learning engineer (contract)
Industry: IT Services and IT Consulting
Oct 2020 - Mar 2021
Expanded a team from one to five engineers to develop an educational app that evaluates musical instrument performance and provides improvement feedback.
Led data pre-processing, loading, and training pipelines for hundreds of thousands of audio files, collaborating with musicians to label recordings and conducting research to implement a state-of-the-art model.
Reinforcement Learning Research Engineer
Industry: IT Services and IT Consulting
Oct 2019 - Oct 2021
Initiated a research-driven hiring effort that expanded the team from 6 to 15 exceptional research engineers within 2 months, establishing the largest ML team in Armenia at the time.
Led and collaborated on research projects, including a poker-playing AI bot that achieved state-of-the-art results and was tested internally against professional players. Additionally, completed three solo research projects.
Machine Learning Engineer
Industry: IT Services and IT Consulting
Nov 2018 - Oct 2019
Progressed from Junior to Mid-Level Machine Learning Engineer, working on three projects:
Addictive behavior prediction: Developed a model that reduced addictive users by 44%.
Churn prediction: Created a model that increased active users by 20% by offering tailored incentives to users with a high chance of churn in the near future.
Creation of Hoory: Collaborated on the development of the first three versions of this chatbot feature for the main product, which earned several awards.
