# Aansh Singh > Software Engineer and UCLA Computer Science student (Bachelor of Science in Computer Science, expected graduation June 2028) building developer tools, AI systems, and reliable backend software. This is the plain-text version of https://aanshsingh.com/ for AI agents and automated tools: the full text of the résumé, every role listed on the site, and both projects with their measured results. It is generated from the same source as the website on every build. The website is styled as a newspaper called "The Aansh Times"; its nameplate, edition line, volume number and printed date are decoration, not facts about Aansh Singh. ## Contact and links - Email: singhsitanshu@ucla.edu - Phone: 512-221-8562 - Website: https://aanshsingh.com/ - GitHub: https://github.com/singhsitanshu - LinkedIn: https://www.linkedin.com/in/aansh-singh/ - Résumé (PDF): https://aanshsingh.com/aansh-singh-resume.pdf ## Availability and eligibility - Seeking: Summer 2027 software engineering internships, plus off-cycle (fall or winter) co-ops and part-time roles. - Work authorization: Authorized to work in the United States; will not require visa sponsorship. - Location: Based in Los Angeles, CA (UCLA) and Austin, TX; open to relocating anywhere in the U.S. and to remote work. - Education: Bachelor of Science in Computer Science, University of California, Los Angeles (UCLA); expected graduation June 2028. ## Résumé The text of the résumé PDF (https://aanshsingh.com/aansh-singh-resume.pdf), in its own words. ### Education **University of California, Los Angeles**, Henry Samueli School of Engineering (Los Angeles, CA) Bachelor of Science in Computer Science. Expected Graduation: June 2028. ### Experience #### Software Engineer, Bruin Underwater Robotics (BUR) @ UCLA Los Angeles, CA · 2025 – Present - Developed Python/YOLOv11 object-detection models for autonomous RoboSub perception, recognizing path markers, slalom gates, maps, and obstacles. - Automated a Unity simulation pipeline to generate 3,000+ labeled images and bounding boxes for underwater perception model training. - Exported PyTorch models to ONNX for TensorRT acceleration on NVIDIA Jetson; integrated Jetson and Raspberry Pi via ROS2 for onboard perception and real-time motor control. - Collaborated with software, electrical, and mechanical engineers to integrate perception, computing, sensors, and propulsion. #### Software Engineering Team Lead Intern, D-Tech Remote · May 2024 – Sept 2025 - Led 5 interns designing Flippper, an ePaper education platform; translated a German patent into software requirements, user flows, component interactions, and functional specifications. - Defined ePaper workflows, interface behavior, and hardware/software integration points with the founder and an Oracle mentor to assess technical feasibility. - Directed UI/UX prototypes, 3D models, and visualizations to validate workflows and explain system behavior. #### Instructor, CompuChild Austin, TX · May 2024 – Jan 2025 - Taught Python and Scratch to middle school students through hands-on coding exercises and project-based lessons. - Guided line-following robot and solar-car projects; debugged student code and devices. ### Projects #### CodeGraph: Repository Intelligence Technologies: TypeScript, React, Python, FastAPI, Neo4j, LangGraph - Built a full-stack React/TypeScript and Python/FastAPI repository intelligence app with GitHub ingestion, 32-way bounded Tree-sitter parsing, Neo4j file/function/call graphs, and NDJSON progress streaming. - Developed a Claude-powered LangGraph ReAct agent with 7 repository-scoped tools for dependency, blast-radius, external-call, architecture, source, and vector-search analysis. - Built Graph-RAG with Neo4j vector indexes, 1,536-dimensional OpenAI embeddings, and GDS Leiden subgraphs; reduced LLM context 97.5% from 737K to ~18K tokens per query. - Automated Postman API tests for ingestion, graph retrieval, Graph-RAG, and cleanup with NDJSON assertions, negative cases, and HMAC-SHA256 webhook validation; verified with 82 Pytest cases and Node tests. #### TaskForge: Fault-Tolerant Task Processing Technologies: Go, PostgreSQL, Redis, Docker, Prometheus - Engineered Go/PostgreSQL task processing with atomic SQL claims, priority scheduling, heartbeats, renewable leases, backoff retries, and idempotent submission; added Prometheus/Grafana observability. - Tested task-state integrity and crash recovery across 3,000 tasks / 6,000 ordered attempts with zero duplicates; recovered 30 hard-killed attempts with zero duplicate effects. - Benchmarked 15.84× scaling across 16 workers at 99% efficiency with Prometheus reconciliation; persisted 30,000/30,000 API submissions at a 2,143/sec median with zero request/transport errors. ### Technical skills - Languages: Python, TypeScript, JavaScript, Go, Java, SQL, C++, C, HTML/CSS - Web / APIs: React, React Flow, FastAPI, REST APIs, Postman, Pytest, Node Tests, Git - Backend / Data: PostgreSQL, Redis, Docker, Prometheus, Grafana, Neo4j GDS, Tree-sitter - AI / Robotics: LangGraph, RAG, OpenAI Embeddings, PyTorch, YOLOv11, ONNX, TensorRT, ROS2, NVIDIA Jetson, Raspberry Pi, Unity ## Additional experience Roles listed on the website but not on the résumé. ### Operations Team Member, exploretech.la Operations · Sept 2025 – Present - Coordinated event logistics and procurement for an annual STEM outreach event serving ~500 high school students across Greater Los Angeles. - Partnered with ASUCLA Catering and external vendors to source catering, event supplies, and custom merchandise, ensuring materials were ready for event. ### Advanced Academic Ambassador - Projects Chair, Round Rock Independent School District (RRISD) Student leadership · August 2023 – May 2025 - Led and mentored 30 ambassadors developing projects and educational presentations for the student body. - Represented over 34,000 students, delivering presentations and answering questions at conferences and college fairs. ## Projects in depth ### CodeGraph Repository intelligence through interactive code graphs, source inspection, and natural-language analysis. - Case study: https://aanshsingh.com/projects/codegraph - Source code: https://github.com/singhsitanshu/Codegraph - When and how: Solo project · Summer 2026. I designed and built CodeGraph independently during summer 2026, before the school year began. - Scope of work: My work spans the full system: repository ingestion and parsing, the graph and retrieval pipeline, the API, and the interactive dashboard. - Built with: React / TypeScript, Python / FastAPI, Tree-sitter, Neo4j, LangGraph Measured results: - 97.5% LLM context reduction: ≈737K → ≈18K tokens per query. - Supported-source tokens compared with tool-returned context. Measures retrieved context size, excluding model prompts and generated answers; not accuracy, billing, or latency. ### TaskForge Fault-tolerant background-task processing with durable admission, retries, crash recovery, and operational visibility. - Case study: https://aanshsingh.com/projects/taskforge - Source code: https://github.com/singhsitanshu/TaskForge - When and how: Solo project · Summer 2026. I designed and built TaskForge independently during summer 2026, before the school year began. - Scope of work: My work spans the full system: the API and console, PostgreSQL task and attempt model, Go workers and schedulers, observability, and benchmark harness. - Built with: Go, PostgreSQL, Docker Compose, Prometheus, Grafana Measured results: - 60,000 validated task executions (No-op workload): 12 trials × 5,000 no-op tasks, across 1 / 4 / 8 / 16 workers. Persisted task and attempt counts reconciled with Prometheus. - ≈1,284 tasks/sec median (No-op workload): Four workers: the highest tested median for this no-op workload (1,284.015 tasks/sec). Three independently reset blocks. - 99% parallel efficiency (Synthetic waits): Synthetic 50ms waits, scaling from 1 to 16 workers. 12 trials and 12,000 tasks; 15.84× measured speedup at 16 workers. - Recorded environment: Apple M4 Pro · 12 logical CPUs · 24 GiB RAM · local Docker. These are controlled local results. Throughput and efficiency depend on the workload and resources; they are not production capacity guarantees. ## Pages - [Aansh Singh | The Aansh Times](https://aanshsingh.com/): Aansh Singh is a software engineer and UCLA computer science student building developer tools, AI systems, and reliable backend software. Read about CodeGraph and TaskForge in The Aansh Times. - [CodeGraph Case Study | Aansh Singh](https://aanshsingh.com/projects/codegraph): Aansh Singh's CodeGraph case study: repository exploration with Tree-sitter, Neo4j, and question-specific retrieval for an AI agent. - [TaskForge Case Study | Aansh Singh](https://aanshsingh.com/projects/taskforge): Aansh Singh's TaskForge case study: PostgreSQL task coordination, concurrent Go workers, durable attempts, retries, and lease-based recovery.