Syed Jibran Haider Email and Phone Number

At YamaSciences, a cutting-edge startup, I am developing an AI-powered data-insights pipeline that automates pharmaceutical marketing content creation & compliance.DATA ACQUISITION & INTEGRATION:• Data Collection — Manually collect, clean, & process Structured Product Labeling (SPL) & clinical trials data.• APIs & Web Scraping — Leverage APIs & web scraping tools (BeautifulSoup, Scrapy, Selenium) for automated data retrieval & integration.GENERATIVE AI & DATA ANALYSIS:• AI Pipelines — Build RAG (Retrieval-Augmented Generation) pipelines using Haystack, integrating Large Language Models (LLMs) for high-fidelity insights across structured & unstructured data.• AI Optimization — Evaluate prompt engineering techniques & pipeline configurations, including document stores, embedders, retrievers, & generators. Experiment with retrieval methods (e.g. dense vs. hybrid) & integrate vector databases like Qdrant.• Data Analysis & NLP — Explore Natural Language Processing (NLP) techniques & experiment with ML algorithms to analyze key pharmaceutical data.VISUALIZATION & INTERACTIVE SOLUTIONS:• Data Visualization — Develop clear & actionable data visualizations using Matplotlib, Seaborn, & Plotly to support strategic decision-making.• AI Chatbot Development — Design an AI-powered chat interface that enhances user interaction with data by providing context-aware responses & intelligent insights.COLLABORATION & STRATEGIC CONTRIBUTION:• Collaboration — Collaborate with cross-functional team to refine data models & contribute to product's strategic direction, ensuring alignment with business objectives & regulatory standards.TOOLS & TECHNOLOGIES:• Programming — Python, OOP• AI Frameworks — Deepset Haystack, OpenAI, Anthropic• Vector Databases — Qdrant• APIs & Web Scraping — RESTful APIs (OpenFDA, ClinicalTrials.gov), BeautifulSoup, Scrapy, Selenium• Data Processing — Pandas, NumPy• ML — Sklearn, TensorFlow, PyTorch• Visualization — Matplotlib, Seaborn, Plotly

- Evaluate and refine AI and machine learning models through completion of over hundreds of analytical tasks across more than 15 distinct categories, enhancing the quality of digital services and products.- Apply detailed analytical skills to a variety of AI training data including text, images, audio, and video, playing a role in the development of multilingual technology platforms.- Interpret and adhere to complex guidelines for task execution, employing rigorous research and evaluation to ensure data integrity, with tasks often necessitating comprehensive written feedback.- Contribute to the accuracy and enrichment of digital map content, utilizing profound geographical knowledge and research acumen to inform global user navigation systems.- Demonstrate flexibility and self-direction as an independent contractor, managing task workflows with fully flexible hours while upholding high standards of quality and consistency in data analysis.

Developed machine learning algorithms to investigate complex cosmological problems as a graduate researcher at CITA. Awarded $105,000 for NSERC CGS-D, a highly competitive national scholarship granted to 330 out of 1,721 applicants, and $30,000 in other scholarships recognizing outstanding academic achievements and research potential.---MACHINE LEARNING PIPELINE DEVELOPMENT:- Explored blind source separation (BSS) algorithms to extract non-Gaussianity from early Universe cosmic fields, pioneering novel signal processing methods in cosmology.- Implemented a multi-scale, Fourier-filtered FastICA algorithm, addressing underdetermined problem domains and localization issues and exploring field statistics.- Selected as 1 of 5 OpenAI GPT-4 Test Pilots by the CITA Director to evaluate advanced AI models for research.DATA MANAGEMENT & SOFTWARE PROFICIENCY:- Leveraged Python libraries (NumPy, SciPy, Scikit-learn, Matplotlib) as well as MATLAB to simulate, process, and visualize complex cosmic signals.- Implemented Python modules to handle and visualize large cosmic simulations, processing 3D cubes with up to 8 billion voxels and simulating 1D fields with over 4 million datapoints.- Processed and visualized ~18 million pixels of astronomical image data in various formats with Python, demonstrating expertise in big-data management.- Developed remotely on high-performance compute (HPC) terminals in Toronto and Oslo, deploying Git and utilizing Jupyter Notebooks to optimize project operations.GLOBAL COLLABORATION:- Collaborated in a global team of 20+ researchers (Cosmoglobe), utilizing effective communication skills to disseminate complex results and contributing to ongoing research on Bayesian component separation of astronomical data (paper in prep).- Participated in international workshops in Santa Barbara (USA) and Oslo (Norway), engaging with the scientific community and enhancing skills in Bayesian analysis, astronomical noise handling, and more.---

Delivered multifaceted instructional support across 7 astronomy/astrophysics courses as a Graduate Teaching Assistant at the University of Toronto. Bolstered engagement and learning for hundreds of students, blending my deep understanding of astrophysics, mathematics, and programming with a strong knack for teaching.---INSTRUCTION & MENTORSHIP:- Led ~50 engaging tutorial sessions, organized 20+ help-sessions/office-hours, and managed classrooms with up to 70 students for a diverse set of courses, fostering active participation and communicating complex concepts effectively.- Consistently commended for ability to simplify scientific concepts and guide students through thought-provoking Socratic questioning.CONTENT CREATION, PROBLEM SOLVING, & TEAM COLLABORATION:- Crafted and contributed to 100+ problem sets for tutorials, exams, and assignments. - Designed Python coding problems in Jupyter Notebooks, bridging the gap between theoretical knowledge and applications.- Collaborated effectively with diverse teams of up to 20 TAs and professors per course to ensure successful course delivery. - Coordinated efforts in creating course content and resolving student queries, demonstrating strong teamwork and leadership skills.COMMUNICATION, SUPPORT, & PRACTICAL APPLICATION:- Graded 1500+ student submissions and managed 100s of email/online queries, providing extensive feedback and delivering robust student support online and in-person.- Adapted to remote teaching methods for 5 courses during the pandemic, showcasing strong digital communication skills and flexibility.- Fostered practical understanding by assisting students with experimental astronomy projects (e.g. telescope operations) in 4 courses, demonstrating technical expertise and ability to guide others in applying theoretical knowledge to practical situations.---

Leveraged deep learning techniques to model and predict the evolution of N-body simulations of galactic dynamics, achieving a significant speedup of over 100x and near-perfect accuracy. Generated and managed large-scale simulations and data pipelines to enable fast and efficient emulations, creating hundreds of data visualizations.---DEEP LEARNING MODEL DESIGN, APPLICATION, & VALIDATION:- Developed a PyTorch-powered Multilayer Perceptron (MLP) model with ~4.5 million parameters to accurately emulate complex N-body simulations.- Optimized the neural network with well-motivated, iterative improvements, ultimately implementing an effective 5-layer architecture with stochastic gradient descent (SGD) as the optimizer, mean-squared error (MSE) as the loss, and Tanh as the activation function.- Achieved a massive speedup of 100-1000x with near-perfect accuracy over traditional N-body simulations, utilizing the Kolmogorov-Smirnov (K-S) divergence test for model accuracy validation.DATA MANAGEMENT & SOFTWARE PROFICIENCY:- Engineered a Python-based pipeline to generate and visualize the dynamics of tens of thousands of simulated galactic discs using: NumPy, SciPy, Matplotlib, GalPy (galactic dynamics library) and wendy (N-body code).- Created substantial datasets of ~10,000 phase-space profile samples to train and test the MLP model.- Managed large-scale N-body simulations, dealing with ~5000 particles over 1000 time-steps each, showcasing proficiency in handling large-scale data.- Employed Git for efficient project collaboration and management.---

Awarded US$4,000 by the National Science Foundation and University of Richmond to conduct research in astrophysics. Developed novel statistical techniques to enhance correlation analyses in large astrophysical datasets (~50,000 datapoints), overcoming limitations of previous methods (Singal et al., 2019) [1]. Actively involved in addressing a significant unsolved astrophysical problem – excess levels of radio synchrotron (Singal et al., 2018) [2].---DATA SCIENCE & STATISTICAL ANALYSIS:- Devised and implemented innovative data binning and analysis strategies (such as using Partial Pearson Correlation statistics) in IDL, a programming language learned from scratch.- Processed ~50,000 quasar measurements in structured datasets.- Synthesized complex results to perform cogent data visualization, generating and studying 100s of correlation analysis iterations.COMMUNICATION, COLLABORATION, & PROJECT MANAGEMENT:- Co-authored 2 journal articles and delivered talks at 3 international conferences [1, 2, 3, 4, 5].- Co-organized the Radio Synchrotron Background Conference (RSBC), coordinating logistical details and actively collaborating with over 20 leading astrophysicists.- Played a pivotal role in writing the summary report for RSBC [2], synthesizing unfamiliar, technical concepts into publication-ready material.- Took initiative as the only undergraduate speaker at MARLAM6 [3] and one of only 2 at RSBC [4], presenting complex ideas to over 40 experts across the 2 meetings.---[1] J. Singal, ..., *J. HAIDER*, et al. (2019). ApJ. 877(1): 63 (13pp). https://doi.org/gqtfnn.[2] J. Singal, *J. HAIDER*, et al. (2018). PASP. 130(985): 036001 (21pp). https://doi.org/gczwht.[3] *J. HAIDER* (Oct 2018). MARLAM6, Baltimore, USA. https://tinyurl.com/2s4txwjm. (Talk).[4] *J. HAIDER* (Jul 2017). RSBC, Richmond, USA. https://tinyurl.com/mr366h5c. (Talk).[5] *J. HAIDER* (Jan 2018). 231st AAS, Washington, USA. https://tinyurl.com/6v356vak. (Poster).---

As a Resident Assistant (RA) employed by the University of Richmond, mentored and supported 135 students over 3 years, empowering them to thrive in academic, social, personal, and professional pursuits.- Fostered community, mediated conflicts, ensured safety and security, strengthened diversity, and worked with multiple teams to organize educational and social events. - Monitored apartments blocks and residence halls at scheduled nights to identify policy infringements and alarming situations, responding appropriately and documenting incident reports.

Undertook mathematical modelling and computational analysis of stochasticity in ultrasensitive molecular biocircuits, earning a US$4,000 research fellowship. Conducted novel biophysics research, using computational tools for model simulation and presenting results at an international conference [1]. Contributed to the understanding of signal flow through stochastic biocircuits of molecular interactions, vital for both systems and synthetic biology.---MATHEMATICAL MODELLING & COMPUTATIONAL PROFICIENCY:- Developed and implemented a mathematical model of stochastic interactions in molecular biocircuits, demonstrating expertise in formulating complex mathematical modelling processes.- Leveraged Mathematica to simulate variances in molecular biocircuits using a form of the Pauli Master equation, displaying proficiency in using computational software to solve intricate scientific problems.RESEARCH & PROBLEM-SOLVING:- Conducted an analysis of stochastic signals produced by bio-circuits containing feedback loops, in particular studying a modified version of the ultrasensitive MAPK (mitogen-activated protein kinase) cascade.- Successfully simulated the modified MAPK bi-stable switch, illustrating a robust understanding of biological systems and the ability to apply mathematical principles to biological phenomena.COMMUNICATION & MENTORING:- Presented research findings at the 83rd SESAPS Conference [1], refining the ability to effectively communicate complex research ideas to a diverse audience.- Mentored a URISE (University of Richmond Integrated Science Experience) student as a voluntary initiative, showcasing leadership and the capacity to foster learning in others.---[1] *J. HAIDER* (Nov 2016). South Eastern Section of the APS Conference, Charlottesville, USA (Poster). https://tinyurl.com/bd8tur9k---

- Guided customers on phone & in person about services offered at the Technology Learning Center, including printing, scanning & software.- Provided instructional technology production & development support to Information Services staff using HTML design skills, advanced authoring tools, multimedia software & peripheral multimedia equipment.- Assisted faculty, staff, & students in using multimedia applications, creating HTML documents or using HTML editing software for instructional-based academic projects

Earned a US$4,000 fellowship to pursue theoretical cosmology research at the renowned Perimeter Institute for Theoretical Physics. Performed mathematical and computational modelling to predict novel cosmological phenomena, synthesizing and presenting findings in an honors thesis and at an international conference [1].---SELF-DIRECTED LEARNING, RESEARCH, & PROBLEM-SOLVING:- Independently assimilated advanced concepts including general relativity, Einstein notation, Feynman path integrals, quantum field theory, and complex analysis through rigorous and rapid self-study.- Analyzed the Schwinger effect using quantum mechanical path integral formalism and Picard-Lefschetz theory, showcasing the ability to synthesize knowledge across different domains (physics and mathematics) to solve intricate problems.- Demonstrated an aptitude for independent, self-driven work and resilience in the face of complex challenges, critical for thriving in the fast-paced tech industry.MATHEMATICAL MODELLING, COMPUTATIONAL PROFICIENCY, & VISUALIZATION:- Deployed Mathematica and Maple to implement mathematical models, solve differential equations of motion, and generate complex visualizations.- Performed numerical approximations to solve oscillatory propagator integrals and plot particle trajectories in Minkowski and de Sitter space, demonstrating proficiency in using computational tools to analyze scientific problems.SCIENTIFIC COMMUNICATION & OUTCOMES:- Presented research findings at the 233rd AAS Meeting [1], engaging with an international audience.- Produced an undergraduate honors thesis, contributing to academic knowledge base and refining abilities to articulate research findings in a professional, clear, and detailed written format using LaTeX.---[1] *J. HAIDER* (Jan 2019). 233rd AAS, Seattle, USA (Poster). https://tinyurl.com/84wue7ye---