a
ahmedtahir_dev

Ahmed S

@ahmedtahir_dev

Software Engineer

Pakistan
Deutsch, Englisch, Urdu
Einige Informationen werden in englischer Sprache angezeigt.
Über mich
I am a software engineer with experience in AI/ML, full-stack development, and data engineering. I have a proven track record of building and deploying scalable solutions, improving user experience, and driving business insights.... Mehr lesen

Kompetenzen

a
ahmedtahir_dev
Ahmed S
offline • 
Durchschnittliche Antwortzeit: 2 Stunden

Meine Dienstleistungen

KI-Technologie-Beratung
I will develop ai agents, chat bot, llm rag, and model finetuning

Portfolio

Arbeitserfahrung

Careem

Software Engineer

Careem • Vollzeit

May 2023 - Apr 20251 yr 11 mos

- End-to-end data collection and processing: Designed and implemented robust pipelines to collect, clean, and transform data for model development. - Development and testing of models: Built, trained, and validated machine learning models to solve real-world business problems. - A/B or A/B/C testing and drive actionable insights: Conducted A/B and multivariate testing to evaluate model performance and translate findings into strategic recommendations. - Deployment of the models and their scaling: Deployed machine learning models to production environments and optimized them for scalable, high-availability usage. • Conducted pre- and post-experiment analysis to measure retention rate uplift, AOV comparison (Treatment vs. Control), and the impact of promos and nudges. • Worked on Food & Grocery tree-based churn models, achieving an AUC of 0.85 and a precision of 80% • Developed and deployed a python service with integrations of OpenaAI, langchain and Langfuse releasing Rest APIs with 82% code maturity and testing level • Automated the JIRA ticket routing process by implementing a 3 models classifier in Bug triage system using OpenAI GPT embeddings, vision and chat API attaining 80% accuracy on the test dataset. • Utilized a Bug Priority Classification model in a two-tier architecture, incorporating Catboost. Integrated a GenAI LLM model for interactive analysis of bug details. This system enhanced the efficiency of bug resolution by prioritizing and categorizing issues, leading to more streamlined development workflows. • Implemented a ConvLSTM-based demand prediction model into production. This model accurately forecasts food order demand for the next 24 hours with hourly granularity, enabling the optimization of resource allocation and supply chain management. • Developed a pseudonymisation script ensuring data privacy before sharing with other parties, completed within 7 days, using Clean Code design principles. • Participated in collaborative academic projects, c