Ahmad is available for hire
Hire AhmadAhmad Abdullah
Verified Expert in Engineering
Data Engineer and Developer
Islamabad, Pakistan
Toptal member since August 17, 2026
Bio
Ahmed is a senior AI/ML and data engineer who co-founded TensorLabs and led it as head of AI, scaling to a 25-engineer team and 80+ clients. Over 7 years, he built Qualzo, an AI-twin platform with 7,000+ active users, and led the AI/ML team at Deep3 Labs, MinersAI, and Mindreader.
Portfolio
CODEplus Inc.
Artificial Intelligence (AI), Agentic AI, Prompt Engineering...
MinersAI
FastAPI, AWS Lambda, Step Functions, GDAL, Isolation Forest, CatBoost, NetCDF...
Tensor Labs
Artificial Intelligence (AI), Machine Learning Operations (MLOps)...
Experience
- Machine Learning - 8 years
- Artificial Intelligence (AI) - 8 years
- ETL - 7 years
- Amazon Web Services (AWS) - 6 years
- Agentic AI - 5 years
- Retrieval-augmented Generation (RAG) - 5 years
- LLM Agents - 5 years
- Microsoft Azure - 4 years
Preferred Environment
Amazon Web Services (AWS), Docker, Microsoft Azure, Artificial Intelligence (AI), Anthropic Claude API, Machine Learning, Machine Learning Operations (MLOps), Agentic AI, LangChain, Retrieval-augmented Generation (RAG)
The most amazing...
...pipeline I've shipped uses LLMs to turn thousands of pages of geological survey reports into structured, cited insights, in production on AWS.
Work Experience
AI Consultant (Contract)
2026 - PRESENT
CODEplus Inc.
- Retained as the sole AI consultant to lead the AI transformation of engineering delivery for a Dublin software consultancy behind a 20-year enterprise platform. Assessed the entire team and designed their AI-assisted development workflow end to end.
- Architected the AI layer of the platform's full rebuild: agent-driven development workflows, AI first-pass code review, and automated QA generation integrated into delivery.
- Drove AI adoption across the full engineering team: 1-on-1 discovery with every developer, all 5 live on tailored AI workflows within the first week, backed by per-developer guides and teach-in sessions.
Technologies: Artificial Intelligence (AI), Agentic AI, Prompt Engineering, Agentic AI Systems, LangSmith, LangGraph, Agentic Frameworks, Agentic RAG Systems, Agentic Workflow Design, AI Agents, LLM Integration, Claude, Claude API, Amazon EC2, AI Agent Orchestration, Vector Databases, RAG Pipelines, RAG Systems, Anthropic, OpenAI, API Development, Machine Learning (ML) APIs, Python 3, Conversational AI, Automation, Claude Agent SDK, ChatGPT, Generative Artificial Intelligence (GenAI), Microsoft Copilot, SQL, Git, Communication, Attention to Detail, AI Systems, Multi-agent Orchestration, Large Language Models (LLMs)
Senior Machine Learning & Data Engineer
2024 - 2026
MinersAI
- Owned the back-end, machine learning, and ETL pipelines of a geospatial mining-exploration SaaS: FastAPI on AWS Lambda with Step Functions state machines orchestrating every upload and computation across three AWS environments.
- Built the LLM layer end to end—an analyst pipeline that reads geological PDF reports into structured observations and summaries, and RAG chat over a document library on an OpenSearch vector store.
- Designed the platform's entire ETL pipeline on ECS Fargate with GDAL for vector and raster geodata, including agentic steps that automatically classify and route incoming datasets.
- Built the ML computation modules: Kriging and IDW heatmap interpolation, Isolation Forest anomaly detection, CatBoost multi-model prospectivity stacking, and NetCDF data cubes.
- Engineered the ingestion path to push 1-gigabyte single-file geodata uploads past AWS Lambda's 15-minute ceiling, producing multi-band cloud-optimized GeoTIFFs in production.
- Removed the platform's Mapbox dependency by engineering an in-house Tippecanoe vector-tile pipeline, with TiTiler serving cloud-optimized GeoTIFFs.
Technologies: FastAPI, AWS Lambda, Step Functions, GDAL, Isolation Forest, CatBoost, NetCDF, Amazon OpenSearch, Amazon Bedrock AgentCore, LangChain, Retrieval-augmented Generation (RAG), GitHub Actions, Architecture, Databases, Data Modeling, Data Governance, Prompt Engineering, AI Engineering, AI Model Training, Software Engineering, Agentic AI Systems, LangSmith, LangGraph, Agentic Frameworks, Agentic RAG Systems, Agentic Workflow Design, AI Agents, LLM Integration, API Design, Pandas, Data Engineering, Tippecanoe, Mapbox, OpenAI API, Model Context Protocol (MCP), Scikit-learn, PostgreSQL, MySQL, Parquet, DuckDB, GeoPandas, Rasterio, Shapely, Alembic, Claude, Claude API, Amazon EC2, Vectors, AI Agent Orchestration, AI Voice Agents, Large Language Model Operations (LLMOps), Vector Databases, Workflow Automation, RAG Pipelines, RAG Systems, Anthropic, OpenAI, PDF Scraping, API Development, Machine Learning (ML) APIs, Python 3, Conversational AI, APIs, AI Architecture, AI Integration, Automation, Geospatial Data, Geospatial Analytics, Startups, Claude Agent SDK, RAG Architecture, Model Evaluation, Streamlit, Visual Language Models (VLMs), Back-end, ChatGPT, Generative Artificial Intelligence (GenAI), Microsoft Copilot, Product Engineering, Cloud Run, Applied AI, Vertex AI, Python, SQL, Artificial Intelligence (AI), Multi-agent Systems, Kimi, Multimodal Models, Git, Azure, Communication, Data Pipelines, Feature Engineering, Model Development, Stakeholder Management, Model Deployment, Code Review, Attention to Detail, Team Leadership, Interviewing, Team Management, Written Communication, AI Systems, Multi-agent Orchestration, Large Language Models (LLMs), Supabase
Co-founder & Fractional Head of AI
2023 - 2026
Tensor Labs
- Co-founded an AI/ML services company and scaled it to 25 engineers, 80+ clients, and $1.1 million in revenue in 3 years.
- Directed delivery across LLM products, RAG pipelines, agentic systems, computer vision, and clinical NLP for clients in the US, EU, Middle East, and APAC.
- Incubated and shipped company products: Shroudly, an AI document-redaction engine for GDPR/HIPAA compliance (OCR ingestion, PII/PHI detection), and RecorDr, a clinical documentation AI.
- Launched Qualzo, an AI-twin platform that turns a person's real data into a RAG-grounded conversational profile; 1,100 active users in its first 12 days with zero ad spend.
- Built the company's AI-assisted engineering practice and its agent-driven development workflows, now sold as a consulting offer.
- Set architecture and hiring bars, ran technical reviews, and personally closed and delivered flagship engagements in healthcare AI and sales intelligence.
Technologies: Artificial Intelligence (AI), Machine Learning Operations (MLOps), Optical Character Recognition (OCR), Large Language Models (LLMs), Retrieval-augmented Generation (RAG), Natural Language Processing (NLP), Computer Vision, Agentic AI, Deep Learning, Amazon Web Services (AWS), GitHub Actions, Architecture, Database Architecture, Databases, Prompt Engineering, Data Analysis, Data Analytics, Analytics, AI Engineering, AI Model Training, Software Engineering, AI Modeling, Agentic AI Systems, LangSmith, LangGraph, Agentic Frameworks, Agentic RAG Systems, Agentic Workflow Design, AI Agents, LLM Integration, API Design, Pandas, Dashboards, Data Engineering, PyTorch, OpenAI API, Model Context Protocol (MCP), Scikit-learn, Keras, PostgreSQL, MySQL, Alembic, Claude, Claude API, Amazon EC2, Vectors, AI Agent Orchestration, AI Voice Agents, Large Language Model Operations (LLMOps), Vector Databases, Workflow Automation, RAG Pipelines, RAG Systems, Anthropic, OpenAI, PDF Scraping, API Development, Machine Learning (ML) APIs, Image Processing, Python 3, Azure AI Document Intelligence, Sentiment Analysis, Conversational AI, APIs, AI Architecture, AI Integration, Google Cloud Platform (GCP), Automation, Predictive Modeling, Data Scientist, Geospatial Data, Geospatial Analytics, CTO, Early-stage Startups, Startups, Technical Leadership, Leadership, New Business Ventures, Claude Agent SDK, RAG Architecture, Healthcare, Model Evaluation, Google Gemma, OpenCV, Streamlit, Visual Language Models (VLMs), Back-end, ChatGPT, BigQuery, Generative Artificial Intelligence (GenAI), Microsoft Copilot, Product Engineering, Cloud Run, Applied AI, Vertex AI, Python, SQL, Multi-agent Systems, Object Detection, Kimi, Multimodal Models, Git, Azure, Communication, Data Pipelines, Feature Engineering, Model Development, Demand Forecasting, Stakeholder Management, Model Deployment, Code Review, Attention to Detail, Team Leadership, Screeners, Interviewing, Team Management, Written Communication, AI Systems, Multi-agent Orchestration, Node.js, Supabase, TypeScript
Lead Data Scientist
2023 - 2024
Mindreader
- Led the entire AI/ML side of the Mindreader application: computer vision, generative AI, MLOps, and the scalable back ends underneath.
- Designed and fine-tuned personality-detection computer vision models (PyTorch, Facetorch) and built their cloud architecture (Docker, Nginx, AWS Step Functions).
- Fine-tuned LLaMA-7B, Falcon, and Mistral 7B for production tasks. Built the SalesGPT back end end to end on LLaMA, MongoDB, and FastAPI.
- Built a RAG-based automation engine that studies client profiles and generates highly personalized cold outreach.
Technologies: Computer Vision, Machine Learning Operations (MLOps), PyTorch, Docker, NGINX, AWS Step Functions, Falcon, MongoDB, FastAPI, LLM Fine-tuning, Architecture, Database Architecture, Databases, Data Modeling, Prompt Engineering, Data Analysis, Data Analytics, Analytics, AI Engineering, AI Model Training, Software Engineering, AI Modeling, Agentic AI Systems, Agentic Frameworks, Agentic Workflow Design, AI Agents, LLM Integration, API Design, Pandas, Dashboards, Data Engineering, OpenAI API, Scikit-learn, TensorFlow, Keras, PostgreSQL, MySQL, Parquet, DuckDB, Claude, Claude API, Amazon EC2, Vectors, AI Agent Orchestration, AI Voice Agents, Large Language Model Operations (LLMOps), Vector Databases, RAG Pipelines, RAG Systems, Anthropic, OpenAI, API Development, Machine Learning (ML) APIs, Image Processing, Python 3, Sentiment Analysis, Conversational AI, APIs, AI Architecture, AI Integration, Google Cloud Platform (GCP), Data Scientist, CTO, Early-stage Startups, Startups, RAG Architecture, Model Evaluation, OpenCV, Streamlit, Back-end, ChatGPT, Generative Artificial Intelligence (GenAI), Applied AI, Python, SQL, Artificial Intelligence (AI), Git, Azure, Communication, Data Pipelines, Feature Engineering, Model Development, Stakeholder Management, Model Deployment, Code Review, Attention to Detail, Team Leadership, Screeners, Interviewing, Team Management, Written Communication, AI Systems, Large Language Models (LLMs)
Lead Machine Learning Engineer
2022 - 2023
GUILDER (Deep3)
- Led ML for Deep3, GUILDER's web3 analytics framework: end-to-end Lifetime Value prediction for Ethereum liquid-staking users and a recommendation system connecting web3 users to their communities.
- Engineered the chain-data ETL and a HODLER scoring model at 92% recall; staking-duration prediction models at 0.86 F1.
- Built GPU-accelerated user-segmentation manifolds with RAPIDS and an initial stakeholder forecasting model from wallet attributes.
Technologies: Ethereum, RAPIDS, Machine Learning, Blockchain, Deep Learning, EDA, Amazon Web Services (AWS), Database Architecture, Databases, Data Modeling, Analytics, AI Engineering, AI Model Training, Time Series, Software Engineering, AI Modeling, Pandas, Dashboards, Data Engineering, Scikit-learn, Keras, PostgreSQL, MySQL, Amazon EC2, Snowflake, AI Agent Orchestration, API Development, OpenAI, Python 3, Sentiment Analysis, APIs, AI Architecture, AI Integration, Predictive Modeling, BigQuery, Applied AI, Python, SQL, Artificial Intelligence (AI), Git, Azure, Communication, Data Pipelines, Feature Engineering, Model Development, Demand Forecasting, Stakeholder Management, Model Deployment, Code Review, Team Leadership, Screeners, Interviewing, Team Management, Written Communication, AI Systems, Large Language Models (LLMs)
Senior Machine Learning Engineer
2021 - 2022
Red Buffer
- Built document entity recognition with LayoutLMv2 (transformers + BERT), combining textual and positional information.
- Delivered COVID-19 case and death forecasting for NIH and sentence-transformer recommendation systems.
- Shipped an end-to-end mobile activity classification pipeline on an ensemble of computer vision models (VGG-16, ResNet, Inception, Faster R-CNN).
- Worked across OCR, GANs, and pattern classification. Ran Flask, RDS, MongoDB, EC2, and Azure functions.
Technologies: LayoutLMv2, BERT, Residual Neural Networks (ResNets), Optical Character Recognition (OCR), GANs, AI Algorithms, Time Series Forecasting, Databases, Data Analytics, Analytics, AI Engineering, AI Model Training, Time Series, Software Engineering, AI Modeling, API Design, Pandas, Dashboards, Data Engineering, PyTorch, Scikit-learn, TensorFlow, Keras, PostgreSQL, MySQL, Amazon EC2, Snowflake, API Development, Machine Learning (ML) APIs, Image Processing, Python 3, Sentiment Analysis, APIs, AI Architecture, AI Integration, Predictive Modeling, Model Evaluation, 3D Pose Estimation, Data Labeling, Gesture Recognition, Google Gemma, OpenCV, Streamlit, Labeling, Pose Estimation, Back-end, Applied AI, Python, SQL, Artificial Intelligence (AI), Object Detection, Hugging Face Transformers, Git, Azure, Communication, Data Pipelines, Feature Engineering, Model Development, Demand Forecasting, Stakeholder Management, Model Deployment, Code Review, Attention to Detail, Team Leadership, Screeners, Interviewing, Team Management, Written Communication
Data Scientist
2021 - 2022
Bird Money (DeFi)
- Designed a deep neural network identifying the top 15% of HODLERs in the Binance ecosystem, and a GRU model predicting wallet trading behavior over 3.4 terabytes of chain data.
- Demoed the wallet holding-time model to Binance Labs; the demo helped bring Binance on board as an investor.
- Ran Binance Smart Chain and Ethereum nodes end to end, extracting raw chain data with ETH-ETL and Web3 tooling into AWS (Athena, Glue, RDS, EC2).
Technologies: GRU, Web3, Machine Learning, ETL, Data Processing, Databases, Data Analytics, Analytics, AI Engineering, AI Model Training, AI Modeling, Pandas, Dashboards, Scikit-learn, TensorFlow, PostgreSQL, MySQL, Amazon EC2, Snowflake, API Development, Python 3, APIs, AI Integration, Data Scientist, Model Evaluation, Python, SQL, Artificial Intelligence (AI), Object Detection, Git, Communication, Feature Engineering, Model Development, Stakeholder Management, Model Deployment, Code Review, Attention to Detail, Interviewing, Written Communication
Data Scientist
2019 - 2021
Human Focus International
- Developed ML pipelines from raw data to AWS deployment through Azure DevOps CI/CD.
- Built the PPE-compliance detection pipeline (CenterNet) and 2 posture-estimation products (AlphaPose, Pose-3D) that significantly increased company revenue.
- Delivered safe-distance detection, driving-behavior analysis, and LSTM/GRU activity detection shipped to CoreML.
- Built in-house chatbot systems with IBM Watson and churn analysis that improved client retention.
Technologies: Azure DevOps CI/CD, Amazon Web Services (AWS), Machine Learning, Deep Learning, ETL, Databases, AI Engineering, AI Model Training, API Design, Pandas, PyTorch, Scikit-learn, TensorFlow, Keras, PostgreSQL, MySQL, Amazon EC2, Image Processing, Python 3, Azure AI Document Intelligence, APIs, Data Scientist, Churn Analysis, Healthcare, 3D Pose Estimation, Data Labeling, Gesture Recognition, OpenCV, Labeling, Pose Estimation, Python, Hugging Face Transformers, Git, Communication, Artificial Intelligence (AI)
Experience
Qualzo
https://qualzo.app/An AI-twin platform that turns a professional's real data (LinkedIn, resume) into a conversational profile visitors can talk to, with every answer grounded in actual source data through RAG rather than free-form model output. I led architecture and delivery end to end: the RAG grounding design, Azure OpenAI integration, and the embeddable chat product. I soft-launched with zero ad spend and reached 1,100 active users in the first 12 days, plus a 200-person waitlist ahead of public launch.
LLM Analyst Pipeline for Geological Survey Reports
For a geospatial mining-exploration SaaS, I built the platform's LLM layer end to end. An analyst pipeline fans out per page over AWS Step Functions to read geological PDF reports into structured, cited observations and summaries, and a RAG chat assistant answers questions over the platform's document library using an OpenSearch vector store. It runs in production across 3 AWS environments, alongside the ETL and ML computation modules I also own: GDAL-based geodata processing, Kriging and IDW interpolation, anomaly detection, and multi-model prospectivity stacking.
AI Extraction Pipeline for Industrial Parts Quoting
https://partfiniti.comPartfiniti is a parts sourcing and quoting platform for the cable assembly and industrial parts industries, and I was the only person on the AI side. I built the pipeline that turns incoming RFQs into validated structured bills of materials. Emails, photographed parts lists, and supplier PDFs that don't follow a shared template go in, and clean, structured data comes out via an API, with a data quality layer that flags bad rows and an enrichment step that fills in missing part information before anything reaches the database. Once the data was structured, the platform generated quotes and orders from it in real time. The work spanned 11 feature deliverables and expanded from a single intake format to the product's back end. The platform has been used to quote over $15 billion in part and cable assembly projects and reports saving over 80% of the time and money the manual process used to take.
Education
2015 - 2019
Bachelor of Science Degree in Computer Science and Data Science
NUCES FAST - Islamabad, Pakistan
Certifications
JULY 2026 - PRESENT
Certified Claude Partner (Claude Code)
Anthropic
AUGUST 2024 - AUGUST 2027
AWS Certified Machine Learning - Specialty
Amazon Web Services
MARCH 2024 - MARCH 2027
AWS Certified Cloud Practitioner
Amazon Web Services
Skills
Libraries/APIs
CatBoost, PyTorch, RAPIDS, OpenAI API, Scikit-learn, TensorFlow, Keras, Pandas, Claude API, API Development, OpenCV, Hugging Face Transformers, Shapely, GDAL, Node.js
Tools
Amazon OpenSearch, AWS Step Functions, Claude Code, Claude, Claude Agent SDK, Visual Language Models (VLMs), ChatGPT, BigQuery, Microsoft Copilot, Git, NGINX, Grafana
Languages
Python, SQL, Python 3, Falcon, Snowflake, TypeScript
Frameworks
LangGraph, Agentic Frameworks, Streamlit, Alembic
Paradigms
Model Context Protocol (MCP), ETL, Automation
Platforms
AWS Lambda, Docker, Amazon Web Services (AWS), LangSmith, Amazon EC2, Google Cloud Platform (GCP), Azure, Ethereum, Mapbox, Blockchain, Cloud Run, Vertex AI
Storage
MongoDB, PostgreSQL, MySQL, Database Architecture, Databases, Data Pipelines
Industry Expertise
Healthcare
Other
Artificial Intelligence (AI), FastAPI, Step Functions, Isolation Forest, Machine Learning Operations (MLOps), Optical Character Recognition (OCR), Large Language Models (LLMs), Retrieval-augmented Generation (RAG), Computer Vision, GRU, Residual Neural Networks (ResNets), GANs, Azure DevOps CI/CD, GitHub Actions, Microsoft Azure, LangChain, Anthropic Claude API, Parquet, Machine Learning, Deep Learning, Agentic AI, Data Science, AI Algorithms, Data Mining, Amazon Bedrock AgentCore, Natural Language Processing (NLP), EDA, Time Series Forecasting, Data Processing, LLM Agents, Architecture, Data Modeling, Prompt Engineering, Data Analysis, Data Analytics, Analytics, AI Engineering, AI Model Training, Time Series, Software Engineering, AI Modeling, Agentic AI Systems, Agentic RAG Systems, Agentic Workflow Design, AI Agents, LLM Integration, API Design, Dashboards, Data Engineering, AI Agent Orchestration, AI Voice Agents, Large Language Model Operations (LLMOps), Vector Databases, Workflow Automation, RAG Pipelines, RAG Systems, Anthropic, OpenAI, PDF Scraping, Machine Learning (ML) APIs, Image Processing, Azure AI Document Intelligence, Sentiment Analysis, Conversational AI, APIs, AI Architecture, AI Integration, Predictive Modeling, Data Scientist, Geospatial Data, Geospatial Analytics, CTO, Early-stage Startups, Startups, Technical Leadership, Leadership, New Business Ventures, RAG Architecture, Model Evaluation, 3D Pose Estimation, Data Labeling, Labeling, Pose Estimation, Back-end, Generative Artificial Intelligence (GenAI), Product Engineering, Applied AI, Object Detection, Multimodal Models, Communication, Feature Engineering, Model Development, Stakeholder Management, Model Deployment, Code Review, Attention to Detail, Team Leadership, Screeners, Interviewing, Team Management, Written Communication, AI Systems, Multi-agent Orchestration, NetCDF, Tippecanoe, Web3, BERT, DuckDB, GeoPandas, Rasterio, LLM Fine-tuning, Data Governance, Churn Analysis, Gesture Recognition, Google Gemma, Multi-agent Systems, Kimi, Demand Forecasting, Supabase, LayoutLMv2, Vectors
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring