Oseremen Ojiefoh

RETICLE (Rationale Engine To Inform CRISPR List Entities) is a web platform that places 2,157 published CRISPR screens on a common calibrated axis so researchers can compare results across studies. Developed with Aaron Gao in the Orvedahl Lab, the project focused on making large-scale CRISPR-screen data more interpretable and comparable.

• Built the front end of the platform — the interfaces researchers use to explore and compare screens on the shared calibrated axis.
• Translated bioinformatics research prototypes into researcher-facing workflows, designing the interfaces around biologists' questions and user stories rather than around the shape of the underlying data.
• Deployed the application to the cloud with CI/CD and single sign-on, turning a local research prototype into an accessible research tool.

A common frustration in math and science education is the gap between a student's written work and meaningful, immediate feedback. Index is an AI-powered mobile and web platform that interprets handwritten STEM work and delivers hint-first, step-by-step guidance rather than just the answer. It is actively being built and tested.

• Shipped an MVP with a custom handwriting canvas, backend services, OCR, and multimodal generative AI.
• Leading hypothesis-driven product testing — defining the critical hypotheses first, then running short customer-discovery cycles to shape MVP features for underserved student segments.
• Evaluating native versus cross-platform architectures for handwriting performance and Apple Pencil support.
• Designing an immersive digital-notebook UI modeled on GoodNotes and Notability, prioritizing low-latency ink rendering and intuitive annotation gestures.
• Early traction: built the first MVP at the Skandalaris Hackathon (Fall 2025), competed in Ideabounce's Open Mic category, and refined customer discovery and go-to-market framing through Olin's entrepreneurship ecosystem and mentorship from the Skandalaris Center.

Every year millions of pets go missing in the United States, yet the process for reuniting them with owners is fragmented across paper flyers, Facebook groups, and disconnected shelter databases. DogTag is a mobile platform designed to close that gap.

• Designed core product flow: photo-based pet reporting, automated geofenced alerts to nearby shelters and volunteers, and a secure in-app messaging layer between owners and rescuers.
• Scraped and normalized public shelter databases to seed a live lost/found index, significantly reducing duplicate manual entry for users.
• Conducted structured customer discovery interviews with 45 pet owners and 20 shelters/rescue volunteers, using findings to prioritize MVP features and de-scope lower-value flows.
• Reached the finalist round of the 2024 Skandalaris Venture Competition, pitching the business model, market size, and technical architecture to a panel of investors and faculty judges.

Built a fully playable browser-based video game controlled entirely through hand gestures, with no physical controller, during a 36-hour hackathon.

• Used MediaPipe Hands to detect and track 21 hand landmarks in real time via webcam, mapping index-finger tip coordinates to in-game movement inputs at ~30 fps.
• Integrated OpenCV for frame capture, preprocessing, and visual feedback overlay, ensuring robust performance under varying lighting conditions.
• Collaborated in a 4-person team, owning the gesture-to-input mapping layer while teammates handled game logic and UI rendering.
• Presented the live prototype to judges; awarded 1st place in the Emerging Track at Hack WashU 2024.

Designed and implemented a closed-loop autonomous driving system on a Raspberry Pi-based vehicle platform from the ground up.

• Designed and tuned real-time PID controllers for motor speed regulation, iteratively adjusting proportional, integral, and derivative gains to minimize overshoot and steady-state error under varying surface friction.
• Integrated DC motor and servo actuators with the control loop for coordinated throttle and steering response.
• Fused data from a photoresistor (ambient light), an ultrasonic sensor (obstacle detection/ranging), and an onboard camera to enable reactive lane-keeping and collision avoidance.
• Validated speed calculations using FFT-based signal analysis, comparing actual against commanded wheel speed to confirm control bandwidth.
• Implemented real-time object detection and color tracking using OpenCV, enabling the vehicle to autonomously follow a designated target at a constant standoff distance.

Designed and built a self-managing raised-bed garden that monitors soil conditions and waters itself without human intervention, powered by solar energy.

• Wired capacitive soil-moisture sensors to Arduino analog inputs and wrote firmware to trigger a solenoid-controlled water valve when moisture fell below configurable thresholds, preventing both under- and over-watering.
• Integrated a solar panel + LiPo charge management circuit to keep the system off-grid indefinitely, including low-power sleep modes between sensor polling cycles to extend battery life.
• Added a DHT22 temperature/humidity sensor and logged readings via serial to a local data store for seasonal analysis of plant health.
• Designed the wiring harness and enclosure for outdoor weather exposure, applying conformal coating to moisture-sensitive boards.

Contributed to a NASA-affiliated research project applying computer vision to classify landscape types in ground-level directional photographs, with the goal of improving GLOBE Observer land-cover classification.

• Fine-tuned a ResNet-18 CNN (PyTorch) on a labeled dataset of sky, land, and water imagery captured across diverse geographic environments, achieving 92–95% test accuracy across the three-class problem.
• Designed and launched a Zooniverse labeling workflow to scale data labeling beyond what the core team could annotate, coordinating volunteer contributors and resolving label conflicts via majority-vote aggregation.
• Analyzed Urban Heat Island and mosquito habitat data using R and ArcGIS.
• Contributed written sections, quantitative figures, and LaTeX formatting to the research paper "Uniting Machine Learning and Citizen Science for Automatic Land Cover Classification."
• Presented findings at the AGU Fall 2023 Meeting (American Geophysical Union) — one of the largest Earth science conferences globally.
• Returned as a team mentor the following cohort, leading project coordination, Discord community management, and supporting new interns through the research workflow.

Independent Stave Company (the world's largest barrel manufacturer) manually counts wooden staves on pallets for inventory and quality control — a slow, error-prone process. I was brought on to automate it with computer vision.

• Evaluated and benchmarked two neural network architectures — CountGD (a grounding-based counting model) and Annolid (an instance segmentation tracker) — for accuracy and inference latency on high-resolution pallet imagery.
• Built the full data pipeline: raw image ingestion, perspective correction, OpenCV-based preprocessing (contrast enhancement, denoising), and structured annotation using LabelMe for model training.
• Trained and evaluated models on an internal dataset of stave pallets, achieving up to 95% accuracy under controlled conditions, and iterating on bounding-box vs. point-based annotation strategies to improve recall on partially occluded staves.
• Developed preprocessing, annotation, and inference workflows for production feasibility, projecting $100K+ in annual savings.

• Epidemiological modeling: Calibrated SIRD and SIRDVB compartmental models in MATLAB using fmincon to fit real COVID-19 case data, then used the calibrated parameters to project the impact of different vaccination rollout strategies and intervention timing on peak case counts.
• Image classification: Built and compared MNIST digit classifiers (logistic regression baseline vs. fully connected neural network), evaluating performance with confusion matrices and per-class precision/recall.
• Unsupervised learning: Applied K-means clustering to county-level COVID trajectory data to identify regions with similar epidemic curves, assessing cluster stability via silhouette scores.
• Signal processing: Designed and analyzed passive RC filter circuits, simulated frequency response, and developed a composite Butterworth low-pass + notch filter pipeline for de-noising biomedical signals with 60 Hz power-line interference.

Explored the integration of AI-powered note-taking into a HoloLens 2 mixed-reality application targeting medical and dental clinical workflows.

• Investigated HoloLens 2 SDK (MRTK3) and spatial mapping APIs to understand feasibility of anchoring UI elements to physical clinical environments.
• Prototyped a voice-to-text and contextual annotation layer using Azure Cognitive Services, allowing clinicians to dictate and tag observations hands-free during procedures.
• Researched LLM integration for auto-summarizing session notes into structured clinical documentation formats.

Enrolled in the McKelvey School of Engineering, with coursework spanning circuits, signals and systems, control theory, and the linear algebra and probability foundations underneath them.

Relevant coursework: Circuits, Signals & Systems, Linear Algebra, Probability, Control Systems.

Activities: Index, Skandalaris Venture Competition, National Society of Black Engineers (NSBE), Minority Association of Premedical Students (MAPS).

GPA: 3.941 UW / 5.007 W  ·  Top 25%  ·  SAT Superscore: 1510  ·  AP Scholar with Distinction

AP courses: AP Computer Science A, AP Calculus BC, AP Physics, AP Chemistry, AP English, AP U.S. History.

Activities: Business Professionals of America (BPA) — Community Service Chair, Java Programming & Financial Portfolio Management competitor; Black Student Alliance; Student Ambassador; Texas Boys State Delegate (2023).

GPA: 4.000 UW / 4.975 W  ·  Class Rank: 26 / 624

Notable coursework: CAD (Autodesk Inventor — earned Certified User credential), Engineering Science, Computer Science II (version control, web design). Transferred to Rick Reedy for junior year.

Completed the competitive NASA SEES internship as a researcher, then returned as a returning mentor to support the next cohort.

• As intern: led the land-cover ML pipeline, co-authored research paper, and presented at AGU Fall 2023.
• As mentor: coordinated team assignments, organized the program Discord server and shared Google Drive, facilitated weekly sync meetings, and guided new interns through dataset preparation and model training workflows.
• Supported junior researchers in preparing their own conference presentations and written reports.

• Served as Community Service Chair, organizing volunteer events and tracking member service hours.
• Competed in Java Programming at the state level.
• Competed in Financial Portfolio Management — placed 5th nationally and reached the top-10 presentation finalist round at the BPA National Leadership Conference.

Selected as a delegate to the American Legion's Texas Boys State program — a week-long civic leadership simulation where students form mock governments and run for elected office.

• Served concurrently as Precinct Chairman, County Chairman, District Co-Chairman, and State Party Delegate.
• Ran a campaign for Commissioner of the General Land Office on the Nationalist Party ticket, delivering speeches and building coalition support among delegates from across Texas.

Co-organized a team, coordinated scheduling and attendance, and led team to finish as the 2nd seed in the division.

Held active roles across three student organizations simultaneously: Community Service Chair (BSA), general member (Multicultural Club), and official Student Ambassador representing the school to prospective students and families during campus visits.

Participated in a community volunteer day at Bonton Farms, an urban-agriculture nonprofit in South Dallas that uses farming to address food insecurity and provide job training in an underserved neighborhood.

• Assisted with produce harvest, sorting, and packaging for community distribution.
• Supported livestock care duties alongside full-time farm staff.