Hi, I'm Eliya Abbas
Software Engineer|
AI Engineer & Data Scientist focused on LLM evaluation, Generative AI and reliable AI systems.
About Me
I'm Eliya Abbas, an AI Engineer and Data Scientist with experience in data analytics, Python, business intelligence and AI-driven application development. I work at the intersection of Machine Learning, Generative AI, LLM Evaluation and software engineering.
As an AI Systems Evaluation Intern at AirDawg Labs, I evaluated LLMs through structured testing, code validation and benchmark-driven assessments (Terminal-Bench, SWE-Bench), using Docker, Linux, Git and WSL. Beyond evaluation, I build AI-powered assistants, automation and full-stack apps with FastAPI, React, Node.js and MongoDB.
I'm particularly interested in building and evaluating intelligent systems that are not only capable, but also reliable, testable, and effective in real-world environments. I'm open to connecting with professionals in AI/ML Engineering, LLM Evaluation, Generative AI, Data Science and AI Systems Development.
Location
Mumbai, India
Education
B.E CSE (AI & ML) — Mumbai University
Focus Areas
LLM Evaluation, GenAI & ML, Full Stack, Data Analytics
Experience
Professional journey, internships and freelance work that shaped my career.
AI Systems Evaluation Intern
AirDawg Labs (Coudal AI)
Evaluated LLM-generated code and responses for correctness and consistency; created and reviewed test cases; worked on Terminal-Bench and SWE-Bench benchmarks; debugged execution failures; used Docker, Linux, Git and WSL for reproducible evaluation workflows; ran prompt-based evaluations and AI model comparisons.
Full-Stack Developer (Freelance)
Self-Employed
Built backend-driven applications with FastAPI and REST architecture, MongoDB-backed API endpoints, JWT authentication, and a modular backend designed for AI workflow integration.
Data Science Intern
The SkyBrisk
Built a content-based Movie Recommendation System (TMDB, 4,800+ titles) with NLTK and cosine similarity, deployed on Streamlit. Ran EDA on 7,000+ Telco churn records with 15+ engineered features, analysed 8,800+ Netflix titles, and automated data preparation to cut manual effort by ~30%.
Business Intelligence Developer / Tableau Developer
Self-Employed
Delivered 100+ interactive Tableau dashboards for a US-based HR and recruitment client (hiring KPIs, attrition, time-to-fill, headcount planning). Built IBM HR Analytics dashboards in Power BI and Tableau and a customer trial conversion and retention dashboard.
Business Analyst (Technical / Web Implementation)
InteliHuman
Built a responsive, SEO-optimized static website with HTML5 and TailwindCSS, loading in under 1 second and delivered production-ready within 8 days.
Business Analyst
Codveda Technologies
Business analysis work in Mumbai, India.
Data Analyst
One IT Tech
Containerized Spark-based data processing jobs with Docker for environment parity across AWS/Azure, and migrated legacy flat-file datasets to cloud storage (S3/GCS) with lifecycle policies and IAM roles.
Data Science Intern
CodSoft
Worked on machine learning classification projects, feature engineering and supervised learning.
Technical Skills
Technologies and tools I work with to build, analyze, and deploy.
Languages & Core
Frontend & Backend
Databases & Cloud
Data & Analytics
AI & Modern Engineering
DevOps & Tools
Reliable, Testable, Measurable
Build it. Test it. Measure it.
I treat AI and data work like software that has to hold up in the real world. I use AI coding tools every day, but nothing ships until I've reviewed it, tested it and checked it against a metric that fits the problem.
Projects
19 projects across AI & ML, data and BI, and full-stack development. Repo-backed projects link to GitHub; client and professional work is described without code.
DeFi Agentic Credit Scorer
End-to-end production pipeline assigning 0–1000 credit scores to Aave V2 wallets using unsupervised KMeans clustering on raw DeFi transaction data for decision intelligence.
On-chain Risk Assessment
Risk scoring engine that rates 100 Ethereum wallets from 0 to 1000 using Compound V2 activity pulled from the Etherscan API, with a weighted, normalised feature model.
Agentic Portfolio Assistant
The "Ask my AI" assistant on this site: answers recruiter questions from my real profile with an LLM on Groq, served by Supabase Edge Functions, with guardrails, rate limits, answer caching, conversation memory and 👍/👎 feedback logged to Postgres.
NLP Recommender Engine
Content-based movie recommender on the TMDB 5000 dataset (4,800+ titles) using NLP feature extraction and cosine similarity, deployed as a Streamlit app with poster integration.
Customer Churn Prediction
Telco churn model on 7,043 customers comparing Logistic Regression and Random Forest on ROC-AUC and recall. Random Forest reached 0.842 ROC-AUC and 77% churn recall; month-to-month contracts churned at 42.7% vs 2.8% on two-year plans.
Flight Delay Prediction
Predicts whether a flight will be delayed from carrier and airport data (Kaggle Airline Delay dataset). Compares Logistic Regression and Random Forest and ships the best model with an evaluation report.
Predictive HR Analytics
Predictive ML model for employee churn analysis, utilizing classification algorithms to drive data-driven retention strategies.
Classification Core Model
Supervised learning implementation for predictive classification, focused on feature engineering and model accuracy optimization.
LLM Code Evaluation & Benchmarking
Professional work · AirDawg LabsEvaluated LLM-generated code and reasoning against Terminal-Bench and SWE-Bench: wrote and reviewed test cases, debugged execution failures, and compared models in reproducible Docker, Linux and WSL environments.
SkyBricks EDA Projects
End-to-end exploratory analysis for my SkyBricks internship: Telco churn data preparation and a Netflix Titles study covering missing-value handling, datetime parsing, feature extraction and trend visualisation.
Decision Intelligence Dashboards
Advanced BI orchestration using Excel, Power BI, and Python to transform complex datasets into actionable decision-making frameworks.
HR & Recruitment Dashboard Suite
Client work · US-based HR firm100+ interactive Tableau dashboards covering hiring KPIs, attrition, time-to-fill, compensation benchmarking and headcount planning, used in weekly executive reviews.
Real-time Service Orchestration
Full-stack QR digital menu platform: Google OAuth and email login, multi-menu dashboard, interactive menu builder, 8 templates, publishing with QR code generation, and a searchable public menu.
This Portfolio Website
Astro site with a Three.js particle hero, light/dark themes, an in-page PDF resume viewer with a highlighter marker, and structured-data SEO.
FastAPI + MongoDB Backend
Freelance workBackend-driven application with REST endpoints on FastAPI and MongoDB, JWT authentication, and a modular architecture built to plug into AI workflows.
InteliHuman Business Website
Client workResponsive, SEO-optimised static site in HTML5 and TailwindCSS with semantic markup and structured metadata. Loads in under a second and was delivered production-ready in 8 days.
Zip Utility App
A faster, redesigned alternative to legacy zip-file tools, built AI-first with an LLM-guided project structure and an SEO-first architecture.
Utility Calculator Site
A set of utility calculators, starting with BMI and mortgage, modernising outdated high-traffic tools as part of a micro-tool portfolio.
Tableau Dashboards
Interactive business intelligence dashboards and data stories built using Tableau Public.
Comprehensive HR dashboard analyzing employee attrition, satisfaction, and demographic trends.
Analytical report on customer trial conversions and behavioral patterns.
Content analysis and viewership trends for Amazon Prime Video.
Deep dive into Netflix content distribution and genre trends.
Comparative analysis of OTT content metrics using Tableau.
GitHub Activity
Real-time stats from my GitHub profile.
I’m a Python Developer and Data Analyst with 2+ years of experience working on real-world data problems, automation, and analytics-driven applications.
Education
Bachelor of Engineering — Computer Science Engineering (AI & ML)
M.H. Saboo Siddik College of Engineering
Mumbai University — Specialized in Artificial Intelligence and Machine Learning
Polytechnic Diploma — Computer Science
Sri Satya Sai University of Technology & Medical Sciences, Sehore
Foundation in computer science fundamentals, programming, and system design
Certifications & Courses
Google Cloud, AWS Academy, Internshala
Perform Foundational Data, ML, and AI Tasks in Google Cloud; Perform Foundational Infrastructure Tasks in Google Cloud; Cloud Computing Fundamentals; Networking; AWS Academy Cloud Foundations; Business Analytics (Internshala); Internship & Job Preparation (Internshala)
Get in Touch
Have a project idea or want to collaborate? Feel free to reach out.
Frequently Asked Questions
Common questions about my background, skills and availability.
Who is Eliya Abbas?
Eliya Abbas is an AI Engineer and Data Scientist based in Mumbai, India, with experience in data analytics, Python, business intelligence and AI-driven application development.
What experience does Eliya have in LLM evaluation and AI benchmarking?
As an AI Systems Evaluation Intern at AirDawg Labs (Coudal AI), Eliya evaluated LLM-generated code and responses for correctness and consistency, created and reviewed test cases, worked with the Terminal-Bench and SWE-Bench benchmarks, debugged execution failures, and ran prompt-based evaluations and model comparisons in Docker, Linux, Git and WSL environments.
What are Eliya's main technical skills?
AI and LLMs: Generative AI, LLM evaluation, AI benchmarking, prompt engineering and machine learning. Programming and backend: Python, SQL, FastAPI, REST APIs, JavaScript, React, Node.js, Express and MongoDB. Data and analytics: Pandas, Power BI, Tableau and Excel. Tools: Docker, Linux, Git and WSL.
What kind of projects has Eliya built?
The portfolio lists 19 projects and 5 Tableau dashboards, including a DeFi credit scoring model using KMeans on Aave V2 wallets, a Telco customer churn model with 0.842 ROC-AUC, an NLP movie recommender deployed on Streamlit, a flight delay prediction model, and a full-stack QR digital menu platform built with Node.js and MongoDB.
Does Eliya have data analytics and business intelligence experience?
Yes. Eliya delivered 100+ interactive Tableau dashboards for a US-based HR and recruitment client covering hiring KPIs, attrition, time-to-fill and headcount planning, and built IBM HR Analytics dashboards in both Power BI and Tableau. Eliya also worked as a Data Science Intern at The SkyBrisk using Python and SQL.
Can Eliya build full-stack and backend applications?
Yes. Eliya built a backend with FastAPI, MongoDB and JWT authentication as a freelance developer, and has built web applications with React, Node.js, Express and MongoDB, including an AI-assisted portfolio chatbot and a QR-based digital menu platform.
What is Eliya's educational background?
Eliya holds a Bachelor of Engineering in Computer Science Engineering (AI & ML) from M.H. Saboo Siddik College of Engineering, Mumbai, and a Diploma in Computer Science. Certifications include Google Cloud (data, ML and AI foundations, infrastructure, cloud computing fundamentals and networking) and AWS Academy Cloud Foundations.
Is Eliya open to new opportunities?
Yes. Eliya is open to opportunities in AI/ML engineering, LLM evaluation, generative AI, data science and AI systems development, and is based in Mumbai, Maharashtra, India.
Which languages does Eliya speak?
English (full professional proficiency), Hindi (native or bilingual) and Marathi (professional working proficiency).
How can I contact Eliya or get the resume?
Email eliyaasayyed@gmail.com, connect on LinkedIn, or view and download the resume.