EXPERIENCE

Ground Software Engineer Intern | Aug 2022 – Nov 2022

Astranis Space Technologies | San Francisco, CA
  • Improved low-latency microservice architecture, reducing satellite noise by 27%, leveraging customer centric mindset in redesign of the Flask backend, React front-end framework, and proprietary stream processor using Elasticsearch within agile team.
  • Proactively identified technical debt and resolving critical issues in Kubernetes deployments, enabling the system to recover and restart several minutes faster potentially preventing multi-million dollar satellite crashes.

Software Developer Intern | Mar 2021 – Nov 2021

Exploratory Systems Laboratory | Davis, CA
  • Created the ResilientBlog and authored the first 6 out of 33 articles on distributed systems, which reduced colleague onboarding time from months to weeks. My efforts were pivotal during the Apache incubation process and resulted in ResilientDB receiving 112 stars and 196 forks — https://go.alejandroarmas.dev/tU0gVJi

Research Scientist Intern | May 2020 - Aug 2020

Massachusetts Institute of Technology - Biology Summer Research Program | Remote, CA
  • Contributed to ML research by deliverying presentation on state-of-the-art reinforcement learning algorithms that learned from reward distributions instead of expected values by reviewing academic literature, proving theorems, and implementing algorithms using Numpy and deep learning frameworks like PyTorch. — https://go.alejandroarmas.dev/UrXJtxv

Intern - Interdisciplinary and Quantitative Biology REU | Jun. 2019 - Aug. 2019

Universidad de Puerto Rico, Rio Piedras | San Juan, Puerto Rico
  • Automated human categorization of bee experiments with k-means clustering solution using R, tidyr, and tibble
  • Communicated weekly to 20+ stakeholders representing 4 teams by creating visualizations with ggplot2. Findings accepted at ABRCMS conference— https://go.alejandroarmas.dev/FUoMsov

PROJECTS

Traffic Prediction Service (Request Access for View) | May 2023 - Present

Tech: Ray, Docker, Terraform, DVC, S3, ONNX, Plotly

  • Streamlined team decision-making by building a cloud experimentation solution, secured by AWS IAM roles/policies, then optimized dataset network transmissions by 35x leveraging DVC, S3, Docker, Terraform, Poetry and Parquet — go.alejandroarmas.dev/qKDhGU2, go.alejandroarmas.dev/LkQpB2H
  • Improved ML model by reducing feature latency from 36 hours to 30-seconds by building a pipeline processing 83K records/s with Apache Flink, then reduced storage by 70% using serialization and compaction.
  • Identified and drove technical requirements insights for an online feature big-data architecture by deeply engaging with unstructured data using GroundingDINO, SAM, DPT, Plotly, OpenCV and CoTracker — https://go.alejandroarmas.dev/RXsjI5F

some-dl-models | Jan 2023 - Apr 2023

Tech: PyTorch, comet ml, Poetry, ONNX, Github Actions

  • Designed and implemented CI/CD pipeline that automated training artifact captures and visualizations saving team ~100 hours of productivity in implementing MLP, CNN, RNN, and GNN models using comet-ml, Poetry, static-analysis, unit-testing, and GitHub Actions — go.alejandroarmas.dev/7hP863H

MLOPS Question + Answer Bot | Mar 2023 - Mar 2023

Tech: Relevence AI, Redis, OpenAI

  • Implemented Vector Search in Q/A Bot and contributed to team being awarded Top 3 at MLOps.community LLM SF Hackathon leveraging Redis and OpenAI APIs — go.alejandroarmas.dev/VstnG18

Automatic Differentiation Backpropagation Engine | Dec. 2021 - Jun. 2022

Tech: C++20, C++17, C++14, OpenCilk, doctest, Catch2

  • Build Deep Learning framework which dynamically builds directed acyclic graphs of tensors and computational operations at runtime. Each node contains a statically built finite state automata, for traversal policies such as gradient descent.
  • Achieved 142x speedup over naive Matrix Multiply and 10x speedup on transpose operations by implementing Parallel Divide and Conquer algorithms and representing matrices in row-major order for reduced cache misses.

D-Cash Wallet Web Server | May 2021 - Jun. 2021

Tech: C++, Stripe API, RapidJSON, POSIX threads, Git

  • Developed secure server wallet authentication, transfer, deposit, and charge issuance services, with integration of Stripe API.
  • Implemented consumer-producer multi-threaded client request concurrent processing scheme to improve latency.

Davis In-Order CPU | Mar 2021 - Jun 2021

Tech: Chisel, Scala, RISC-V, Python, Pandas and Matplotlib

  • Developed a comprehensive RV32I Pipelined CPU design using Chisel HDL, incorporating several functional units such as Branch Predictor, Hazard Unit, Forwarding Unit, ALU, ALU Control Unit, and Control Unit.
  • Conducted extensive simulations and analysis comparing the performance of single-cycle and pipelined designs. Utilized four different branch predictors and varied saturation bits over six workloads to create detailed visualizations.

Lineage-based Data Store | Jan 2021 - Apr 2021

Tech: Python, Git

  • Designed and drove team project to implement cutting edge relational database from academic literature. Built in-memory record indexing and in-memory management, resulting in a Top 5 performance using Python — https://go.alejandroarmas.dev/E7wKir1
  • Mediated system design for concurrency control, contention-free lazy record consolidation and data durability.

TECHNICAL SKILLS

Relevant Coursework:

Skills: C/C++20, Python, Rust, R, Risc-V, JavaScript, SQL (sqlite, postgres), \(\LaTeX\) , CSS, Markdown, HTML, Bash, Spanish Developer Tools: Kubernetes, Docker, Terraform, S3, EC2, Git, Azure, Github Actions, Github Codespaces, Ably

Libraries: NumPy, PyTorch, Pandas, Matplotlib, OpenCilk, doctest, Catch2, RapidJSON, Acme, Reverb, Gym

Honors and Achievements

Professional Memberships:

  • Scholar, NSF LSAMP/CAMP program at UC Davis (June 2021 - 2022)
  • Scholar, AvenueE - VIP Research Program (June 2021 - 2022)
  • Member, Artificial Intelligence Society at UC Davis (Jan 2020 - Jan 2021)
  • Member, NSF LSAMP/CAMP program at UC Davis (Dec 2019 - June 2021)
  • Member, UC Davis AvenueE 2019 Cohort (1 of 30 students) (Sept 2019 - 2022)
  • Member, Institute of Electrical and Electronics Engineers (IEEE)

Awards, Grants and Fellowships:

  • UC Davis ECE Undergraduate Travel Grant ($500)
  • AvenueE Travel Grant ($500)
  • ABRCMS Student Travel Award ($685.57)
  • UC Davis AvenueE Scholarship ($10500)
  • EOPS Perpetual Spirit Scholarship ($1000)
  • LMC Academic Competition scholarship ($500)
  • AvenueE Scholar Award ($1000)
  • LMC Academic Honors Spring 2019
  • LMC Dean’s List 2017-2019