Cody Collier Email & Phone Number

Texas, United States

Richardson, Texas, Us

Working at the intersection of computation and neuroscience in the Price Lab at the Center for Advanced Pain Studies (CAPS)tech: transcriptomics, machine learning, neurobiology

Led search & rag efforts from near 0 to v1 for the core enterprise search productStarted by completing an in-depth comparative analysis of Elastic, OpenSearch, and Vespa to determine the initial search engine to be used by the product. This included a 10+ page artifact to illustrate the analysis and decision.Developed and maintained the core Vespa document schemas to represent consolidated data from across a variety of plugins such as google docs, github, slack, and so on. Collaborated with the team to normalize, map, and align data toward the common schema structures.Developed and maintained the Vespa document processor code responsible for the final stage of processing incoming content to be indexed (from the ingest pipeline). Additionally integrated Embedder functionality to support complex field compositions for semantic embedding.Created most queries and all ranking profiles. Used simple, hand crafted data sets to perform relevance evaluation using NDCG, including a basic evaluation runner. Designed and led the implementation of the CTO's idea for a proprietary, collaboration-based ranking algorithm. Took ownership of the core Vespa searcher with proprietary logic, performed multiple rounds of complexity analysis, and made order of magnitude performance improvements in collaboration with the team.Implemented updates in terraform for Vespa clusters on Kubernetes, with persistent volume mapping and data preservation assurances for both AWS and Azure.Designed and wrote middleware code to produce RAG (Question Answering) prototypes, fully integrated with the core search engine. Led analysis and adoption of multiple LLM integrations.Kept up with industry advancements for AI, LLM, and embeddings, and provided insights and feedback to the team as needed. Supported sales on customer exploration calls with direct customer interactions.Worked as sole search and machine learning SME for the majority of the tenure.tech: Vespa, java, python, golang

Plano, Texas, Us

project: Virtual Assistant - 2022 Lexus NX & TundraFormal Roles:Sr SWE/Architect -> Sr Machine Learning Engineer -> Principal Architect & Managing EngineerMy contributions at TC were varied across software development, architecture, and machine learning. Over time I worked with a team of high quality engineers and product owners to evolve the design of the platform from prototype to production oriented delivery. The core pieces were gol@ng microservices alongside tensorflow models for inference, all arranged in a domain oriented design.I began as a sr swe and architect, focused on breaking apart a monolithic python prototype and helping to design and build microservices. Later we migrated to gol@ng.Early on I was asked to lead a small team with the goal of designing and specifying the communication protocol between a vehicle and cloud services for the project. We strived to meet low latency and robustness goals, and worked mostly with gRPC and protobuf. This spec became the backbone for the project. It involved working in an environment of high ambiguity with several Toyota companies and vendors, as well as collaborating closely with our product owners.In May 2019, I completed my masters degree in computer science and machine learning. Soon after, I made a lateral move to the ml team. There I worked to apply pre-existing classification and ner patterns to one of the NLP domains of the project. This included training and integrating the deep learning models for version one of that domain. Additionally, I completed the first phase of migrating an internal python library from TensorFlow 1 to TensorFlow 2.In my final phase at TC, I returned to platform architecture responsibilities and software team leadership. We worked to deliver the platform for various vehicle program goals. This time included guiding and managing staff, and working closely with our chief product owner and director of engineering.tech: python, tensorflow, NLP, gol@ng, gRPC, protobuf

Denver, Co, Us

My time at Catalyst allowed me to progress through several roles including Platform Engineer in Test, Platform Architect (search), Team Lead, and finally Director of Engineering Research. I was responsible for the engineering aspects of our applied machine learning platform and managed a small team of two senior software engineers.Highlights included leading the engineering side of the backend system which supports the Insight Predict product, as well as conceiving and implementing a custom graph database for the platform. The overall system is responsible for indexing collections of documents, building models designed by our scientist, and leveraging those models for ranking documents. Readers can think of this as a simplified Pandora for documents and e-discovery.Additional technical accomplishments:. Created custom service latency measurement and charting system which was adopted throughout the company to understand platform performance and characteristics (python, influxdb, grafana). Created custom python library and text extraction tool leveraging third-party document library. Created automated testing system for a core HTTP API service of our flagship product, which is still being leveraged and enhanced by other engineers 8 years later (python, unittest). Created python daemon to post/ingest documents into document database clusters. Created a variety of other web and automation tools (php, bash, python, flask)tech: Python, Bash, Linux, RabbitMQ, Redis, Elastic, SwiftStack, Flask, MarkLogic, Machine Learning, proprietary algorithmsThis position was 100% telecommute / remote.

Mountain View, California, Us

As a Udacity Project Reviewer, I reviewed student project submissions, striving to give positive, actionable, and helpful feedback. I reviewed projects for the new Deep Learning Nanodegree program.

Minneapolis, Mn, Us

My core responsibility at RSD was as a senior developer, resolving bugs, making modifications, and adding new features to the flagship product. This product was a web based POS system built upon a stack of php, javascript, mysql, apache, and windows (wamp). Additionally, I worked as a lead developer for a project in which we made custom modifications and feature additions to the product for a high profile customer. This responsibility included technical leadership, work assignment, and management of customer communications.After consistent performance and assistance with process improvement, I was promoted to a development manager. Specifically I was chosen to lead the implementation of automated testing infrastructure (php, selenium) and managed a small team through initial rollout.This position was 100% telecommute.

Denver, Co, Us

At Catalyst (formerly CaseShare) I started as an Activation Technician. My work was initially focused on the front end of the customer life cycle, trying to improve the customer on-boarding process. I worked with peers to try and build consensus and gain adoption of business processes which were co-developed. Though I had little success, I began to learn about influence, message crafting, and process innovation.Later my focus shifted to the initial rollout and implementation of the FAST Instream search platform as a critical component of the company's document review application. As a part of this project, I served as technical owner of the FAST document processing pipeline and related development (python, bash). Additionally, I worked on a variety of other projects including billing report automation and network upgrades.

Palo Alto, Ca, Us

For this contract, I worked on installation and management of remote, VPN connected linux server appliances. These systems were used for monitoring and management of StorageTek devices at international customer sites.

Greenwood Village, Colorado, Us

My work at Viawest began in the network operations center as a NOC Support Technician and eventually progressed to a Tier 2 support role. We monitored a 24/7 four state network of dedicated ISP services, colocation and hosting customers, and data center facilities. We coordinated troubleshooting, internal and external escalation, and event documentation for network outages. Finally, we were responsible for customer support and supervision of escalated issues from tier 1 staff.After a little more than a year, I moved to the hosting group as a Hosting Administrator and later progressed to Sr. Hosting Administrator. My core responsibility was for Linux managed hosting servers and infrastructure (approx. 125 physical servers, 150+ virtual servers). This work included systems administration, systems automation, and customer facing professional services. I worked within a great team to help deliver and manage several successive rounds of hosting products.Finally, I completed my time at Viawest with the additional responsibilities of Hosting Production Supervisor. In this role I managed a small team. We worked to improve the processes for our server supply chain as well as the automated provisioning for Linux and Windows servers.

Lan and web development consulting

Technology generalist and web developer

Master Of Science (M.S.), Computer Science

Bachelor Of Science (B.S.), Cognitive Science