AI Engineering · Technical Program Management

Ibrahim Allahbukhsh

Junior Project Analyst at Citi Treasury

I build AI systems and lead the teams that ship them. At Citi I manage software delivery on an enterprise trading platform and advise leadership on where generative AI can replace manual work — then build the tooling to prove it out.

View my work Get in touch

Featured Project

Financial Research Agent

An agentic AI system that turns live SEC filings into institutional-grade analyst briefs — six tools orchestrated in parallel through LangGraph, every figure cited to its primary source.

LangGraphClaude APIpgvectorRAGFastAPIDocker

Live demo View code

Focus

What I work on

Applied AI

Designing agentic workflows and LLM-backed systems — prompt engineering, RAG, tool orchestration, and evaluation. I advise leadership on where generative AI can replace manual work, then help build it.

Engineering Delivery

Leading a software team enhancing an enterprise trading platform — owning requirements through delivery, from UI changes to new data feeds shipping across DEV, UAT, and PROD.

Data & Automation

Building Python pipelines and reporting that turn manual, hours-long processes into fast, reliable systems — reconciliation tooling, validation checks, and production dashboards.

Toolkit

Technologies

Python

JavaScript

React

Next.js

Node.js

HTML

CSS

Tailwind

Pandas

MySQL

Docker

Git

LangGraphLangChainClaude APIFastAPIPostgreSQLpgvectorOracleSQLTableau

Selected Work

Projects

Financial Research Agent

Agentic AI system that generates institutional-grade analyst briefs using live SEC filings, pgvector RAG, and multi-tool orchestration. Six tools run in parallel via LangGraph with primary-source citations — no fabricated figures.

LangGraphClaude APIpgvectorRAGFastAPIDocker

Live Demo Code

Personal Development Video Analyzer

Custom action extraction module integrating ASR transcription and NLP-based segmentation for video content analysis.

ReactJavaScriptPythonFlaskGemini APIPrompt Engineering

Code

Professor Recommender App

Web application for college students to rate professors and receive recommendations based on their learning styles.

ReactJavaScriptMySQLPHP

Code

Trek

Social platform for hikers to share experiences and plan hiking trips with interactive galleries and trip planning features.

ReactJavaScriptSwagger

Code

News Analysis

News analysis tool using NLP to detect biases and analyze sentiment in articles, helping users make informed decisions.

ReactJavaScriptPythonPostgreSQLFlask

Code

Stock Trade Recommendation

Stock prediction app using reinforcement learning with a Deep Q-Network, achieving 12% ROI on real trades.

PythonReinforcement LearningDeep Q-Network

Code

Career

Experience

Summit Execution Junior Project Analyst

Citi

March 2025 — Present

Used data analytics to support various stages of projects.
Supported user acceptance testing (UAT) to validate that business changes were successfully developed.
Tracked general progress, risks, and issues on a daily basis.
Followed up with project participants to ensure action items from meetings were completed.

Beats by Dre Consumer Insights Data Analytics Extern

Extern

Jul 2024 — Sep 2024

Developed proficiency in Python and various data science libraries to perform sentiment analysis on 1500+ consumer reviews across 17 competitor products.
Applied exploratory data analysis (EDA) techniques to uncover 6 underlying patterns and trends in large datasets.
Utilized advanced natural language processing tools to interpret and summarize 1500+ consumer reviews.

IT Consultant

University at Buffalo

Feb 2022 — Jun 2024

Resolved 100+ technical issues weekly — including software and networking problems — ensuring swift and effective support.
Delivered a range of IT services such as diagnostics, malware removal, system resets, and reinstallation in both Windows and Mac environments.
Streamlined issue resolution through Remedy Force, reducing average ticket response time by 30% for students, faculty, and staff.
Created and improved technical documentation to enhance the efficiency of support procedures.

Data Annotation Specialist

Tesla

Jul 2022 — Aug 2022

Used an autopilot labeling interface to train deep neural networks by accurately labeling hundreds of images.
Coordinated with computer vision engineers to optimize labeling efficiency, contributing interface-design feedback to 2 supervisors.
Deepened understanding of how labels drive learning algorithms, helping design a 98% accurate machine learning model.

Get in touch

Let’s build something.

Open to new opportunities and collaborations. The fastest way to reach me is by email.

Say hello