# Jia Wu

![Jia Wu](/resources/cognition/portraits/jia-rong-wu.webp)

Team Lead, Deployed Engineering at Cognition (via Windsurf/Codeium); ex-Tyk Solutions Architect; Waterloo proteomics-ML researcher  
Cognition  
United States

## At Cognition

- Role: Team Lead, Deployed Engineering (Engineering)
- Joined: Mar 2025
- How they joined: With Windsurf (https://www.autumn.ai/resources/cognition/acquisitions/windsurf)
- Before Windsurf: Solutions Architect, Tyk
- Overlapped before Cognition with other Cognition people at: Codeium (Windsurf) (https://www.autumn.ai/resources/cognition/worked-together)

## Identity
- LinkedIn: https://www.linkedin.com/in/jia-rong-wu
- GitHub: https://github.com/JRWu
- Website: https://jiarongwu.com/
- Domain: cognition.ai

## Notes
- Named '#1 Customer Engineer Revenue Driver of the Year' at Cognition's Sales Kickoff, and wrote: 'Earning this at Cognition hits different.' Thanked Alex Laubscher (day-to-day leadership), Graham Moreno ('setting the standard across Majors') and Nicholas Erdenberger (leading the org). (https://www.linkedin.com/posts/jia-rong-wu_saleskickoff-teamwins-windsurf-activity-7431864630689308673-HTti)
- Spoke at AI Engineer World's Fair 2026 on 'How Forward Deployed Engineering is done at Cognition' (17:38 talk, published 2026-07-28). (https://www.youtube.com/watch?v=RVxym6mmIns)
- Press coverage of the talk reports outcomes Wu cited from embedded FDE work: a 150% increase in effective engineering headcount, an 82% reduction in delivery timelines, and double the merged PRs. Named engagements were Nubank and Build, including a 50-engineer ETL migration done in a third of the time and COBOL/JCL legacy migrations. (https://finance.biggo.com/news/a5f316b781abb2dc)
- Joined Cognition through its July 2025 acquisition of Windsurf (formerly Codeium); ai.engineer says Wu 'worked in deployed engineering at Codeium and Windsurf before joining Cognition'. (https://ai.engineer/speakers/jia-wu)
- Current role: Team Lead, Deployed Engineering at Cognition (2025-present), following a Deployed Engineer role at Windsurf in 2025. (https://www.linkedin.com/in/jia-rong-wu)
- Reshared Cognition's raise announcement ('over $2B at a $48B valuation', per the post URL) with 'job's not finished' and a #hiring call. (https://www.linkedin.com/posts/jia-rong-wu_cognition-has-raised-over-2b-at-a-48b-valuation-activity-7503137877128531968-upYf)
- In the talk Wu says Cognition shipped almost an order of magnitude more PRs internally over six months while hiring lagged, that writing code is usually only about 20% of the problem, and describes a day of 4-5 hours of customer calls followed by 4-5 hours at the keyboard. (https://ai.engineer/talks/RVxym6mmIns-forward-deployed-engineering-is-done-at-cognition)
- Announced the AI Engineer World's Fair talk with: 'At Cognition, I've seen how forward deployed engineering turns frontier AI into measurable outcomes for some of the world's most complex and highly regulated organizations.' (https://www.linkedin.com/posts/jia-rong-wu_forward-deployed-engineering-is-the-most-activity-7476387921256730624-hQbn)
- On the Codeium-to-Cognition arc Wu posted: 'From Codeium -> Windsurf ∪ Cognition, we're redefining how software gets built.' (https://www.linkedin.com/posts/jia-rong-wu_more-devins-in-more-places-cognition-activity-7465646027329613824-ZLw_)
- Published .windsurf, a GitHub repo of sample Rules and Workflows for Windsurf (2025-05), and mcp-sse, a FastMCP 2.0 MCP SSE server in Python (2025-07). Both are customer-facing deployment artifacts from the Codeium/Windsurf period. (https://github.com/JRWu?tab=repositories)
- The 2024 CV credits Wu's solutions role with generating $2M+ ARR through technical consultations, leading on-sites, contributing PRs to the API gateway, and helping move Tyk from Visionary to Leader in Gartner's Magic Quadrant in 2023. (https://github.com/JRWu/cv/raw/main/JiaRongWu_CV_2024.pdf)
- At Tyk Technologies (Solutions Architect, 2021-2025), Wu authored TykTechnologies/custom-go-plugin, a Makefile-based repo for compiling and bundling custom Go middleware. Wu wrote that it 'increased the velocity of customer evaluations' and unblocked deals. (https://www.jiarongwu.com/blog/tyk-custom-go-plugin)
- Co-Founder & Chief Data Scientist at AUSPRE (AUtomatic SPeech REcognition), 2019-2021. Per the CV, Wu trained Transformer models for speaker diarization and speech-to-text in TensorFlow, built ETL pipelines, and won funding through pitches and grants. (https://github.com/JRWu/cv/raw/main/JiaRongWu_CV_2024.pdf)
- Reshared CEO Scott Wu's post on measuring Devin value in productive engineering hours and the up-to-$10M AI productivity guarantee. (https://www.linkedin.com/posts/scott-wu-8b94ab96_measuring-someones-productivity-by-their-activity-7468370692036526081-oB2k)
- Marked the Windsurf rebrand with the post 'The final Wave: Windsurf is now Devin Desktop!' (https://www.linkedin.com/posts/jia-rong-wu_introducing-devin-desktop-the-next-generation-activity-7467748343490949120-vWnc)
- Presented on API Platform Maturity to an audience of over 2,000 technology leaders at Gartner's Application Innovation & Business Solutions Summit 2024, representing Tyk. (https://www.jiarongwu.com/blog/gartner-aibs-2024)
- At the University of Waterloo, Wu built DIA Data Reading, a peptide identification and quantification suite, and trained a CNN to score peptide-spectrum matches on ProteomeXchange data with DeBruijn-generated decoys. The 3D visualizer is open-sourced as JRWu/dia_visualization. (https://www.jiarongwu.com/blog/dia-data-reading)
- Public headline says "Deployed Engineer @ Cognition | Windsurf". (https://www.linkedin.com/in/jia-rong-wu)
- Research interests listed: compositional data analysis for high-throughput omics, peptide identification for tandem MS, and ML for spectrum prediction. TA for CS135, CS231, CS246, CS341, CS482 and CS846 at Waterloo. (https://www.jiarongwu.com/about)
- Was Full-stack Developer (2017-2019) and then Entrepreneur in Residence (2019-2021) at Eighty8 Ventures, building Python/PHP/MySQL microservices with ReactJS. AUSPRE overlaps the EIR period. (https://www.linkedin.com/in/jia-rong-wu)
- Solutions architect by trade, professional academic by training. Interested in all Tech things from research to industry; this comes from the profile summary and preserves the person's own positioning. (https://www.linkedin.com/in/jia-rong-wu)
- Later public repos and forks include invoice-application (JavaScript, 2026-01) and forks of backstage (2025-09) and webpack (2025-09). (https://github.com/JRWu?tab=repositories)
- Blogged technical Tyk material: 'Tech at Tyk' (2023-01) and 'TLS and Tyk' (2023-02), covering Kubernetes, plugins and host-based routing. (https://www.jiarongwu.com/blog/tech-at-tyk)
- At Western University (2015-2016), Wu worked as a software developer refactoring data-analysis pipelines from Perl to C. Open-source JRWu/Log-Ratio-Publication holds scripts for a log-ratio (compositional data) paper. (https://github.com/JRWu/cv/raw/main/JiaRongWu_CV_2024.pdf)
- Previous roles include Solutions Architect at Tyk; Scientific Researcher at University of Waterloo; Teachers Assistant at University of Waterloo; Co-Founder & Chief Data Scientist at Auspre; Entrepreneur In Residence at Eighty8 Ventures; Full-stack Developer at Eighty8 Ventures; Data Analyst & Software Developer at Western University. (https://www.linkedin.com/in/jia-rong-wu)
- Location is United States. (https://www.linkedin.com/in/jia-rong-wu)
- Hobbies listed on the personal site: action cameras, snowboarding and woodworking. (https://www.jiarongwu.com/about)

## Experience
- Team Lead, Deployed Engineering at Cognition - Mountain View, California, United States (Aug 2025 - Present)
- Deployed Engineer at Windsurf - Mountain View, California, United States (Mar 2025 - Present): Windsurf was acquired by Cognition in July 2025 - https://cognition.ai/blog/windsurf

Empowering everyone to dream bigger.

Solutionize, support, and close strategic deals.
- Solutions Architect at Tyk - Americas (Sep 2021 - Mar 2025): Solutionize, support, and close strategic deals in complex sales cycles.

Provide technical guidance and consultation for B2B customers interested in adapting an API-first approach to their business strategy. 

Conduct technical demos, as well as provide strategic guidance on how API management fits into companies, from startups to enterprises. 

Generate, produce and edit content across different media such as videos, blogs and technical articles to help drive both MQLs and SQLs.
- Scientific Researcher at University of Waterloo - Waterloo, Ontario (Sep 2018 - Aug 2022): Project manager responsible for overseeing the software development efforts of the lab. Provided DevOps for lab servers and deployed research applications to on-premise data centres and cloud based Kubernetes clusters. 

Conducted bioinformatics research as a PhD (doctoral) candidate. Built and leveraged machine learning models to identify and quantify proteins from mass spectrometry data. Applied inference to datasets to identify nanoparticle drug candidates utilized in cancer research. Presented at research conferences.
- Teachers Assistant at University of Waterloo - Waterloo (Sep 2016 - Aug 2022): Performed duties as a teaching assistant for the following courses: Object Oriented Programming, Algorithms, Design Patterns, Biological Sequence Analysis, and Functional Programming.

Responsible for creating marking schemes, building automated software evaluation pipelines and performed grading for assignments and exams. 

Consulted with students to assist with their education across multiple computer science disciplines.
- Co-Founder & Chief Data Scientist at Auspre - London, Ontario, Canada (Feb 2019 - Sep 2021): Co-Founded a startup which leveraged AI to perform speech to text and speaker diarization operations. Secured funding from accelerators and led the data efforts of the company. 

Designed, trained, tested and deployed in-house transformer models for the purposes of Speech-To-Text, Speaker Diarization and Speech Refinement. Leveraged Tensorflow and CUDA Toolkit libraries to achieve these goals. 

Built ETL (Extract-Transfer-Load) data pipelines as well as the APIs required for supporting the software.
- Entrepreneur In Residence at Eighty8 Ventures (Feb 2019 - Sep 2021): - Represented the interests of Auspre/Scribr in partnership with Eighty8 Ventures
- Artificial Intelligence domain expert
- Full-stack Developer at Eighty8 Ventures - London, ON (Jan 2017 - Feb 2019): - Configured & maintained a High-Availability Kubernetes Cluster
- Wrote APIs to interface with Amazon AWS, and Google GCP Services
- Built & Deployed Applications on Digital Ocean through Rancher 2.0
- Maintained application data integrity using framework specific ORMs and MySQL
- Data Analyst & Software Developer at Western University - Medical Sciences Building - Gloor Lab (May 2015 - Jun 2016): Developed data processing pipelines for genomic data. Improved the runtime performance of existing pipelines by over 200%, reduced memory footprint of workflows by over 25%. 

Applied statistical analysis to high-throughput sequencing datasets to obtain experimental results utilized in publications. 

Published articles in peer reviewed journals and books.

## Education
- University of Waterloo - Doctor of Philosophy - PhD Candidate, Computer Science - (Jan 2018 - Jan 2022)
- University of Waterloo - Master’s Degree, Computer Science - (Jan 2016 - Jan 2018)
- The University of Western Ontario - Bachelor’s Degree, Honours BMSc Computational Biochemistry - (Jan 2012 - Jan 2016)
- Western University - Bachelor’s Degree, Computer Science - (Jan 2012 - Jan 2016)
- Shad Valley At Memorial University - Shad Valley Certificate of Completion, Business/Mathematics - (Jan 2011 - Jan 2011): - Interacted with professors and attended lectures for the duration of a month
- Led a team of 8 individuals to design and produce a product aimed at "Improving the lifestyle of a Canadian child with a disability."
- Participated in the National Research In Motion/Shad Entrepreneurship Cup as chief web designer winning 3rd place

## Sources
- https://www.linkedin.com/in/jia-rong-wu
- https://ai.engineer/speakers/jia-wu
- https://www.linkedin.com/in/jia-rong-wu?miniProfileUrn=urn%3Ali%3Afsd_profile%3AACoAAA1EBf4BMoLQwGmYoS8TLLqdz8ynqb7y_KI
- https://github.com/JRWu/cv/raw/main/JiaRongWu_CV_2024.pdf
- https://finance.biggo.com/news/a5f316b781abb2dc
- https://github.com/JRWu
- https://github.com/JRWu?tab=repositories
- https://ai.engineer/talks/RVxym6mmIns-forward-deployed-engineering-is-done-at-cognition
- https://www.linkedin.com/in/scott-wu-8b94ab96?miniProfileUrn=urn%3Ali%3Afsd_profile%3AACoAABRcyU0B3eVfSh_wWQgM7w3CZRywJOf28G8
- https://www.jiarongwu.com/blog/dia-data-reading
- https://www.jiarongwu.com/blog/tyk-custom-go-plugin
- https://www.youtube.com/watch?v=RVxym6mmIns
- https://www.jiarongwu.com/blog/gartner-aibs-2024
- https://www.jiarongwu.com/about
- https://www.jiarongwu.com/blog/tech-at-tyk
- https://jiarongwu.com/
- https://www.linkedin.com/posts/jia-rong-wu_saleskickoff-teamwins-windsurf-activity-7431864630689308673-HTti
- https://www.linkedin.com/posts/jia-rong-wu_cognition-has-raised-over-2b-at-a-48b-valuation-activity-7503137877128531968-upYf
- https://www.linkedin.com/posts/jia-rong-wu_forward-deployed-engineering-is-the-most-activity-7476387921256730624-hQbn
- https://www.linkedin.com/posts/jia-rong-wu_more-devins-in-more-places-cognition-activity-7465646027329613824-ZLw_
- https://www.linkedin.com/posts/scott-wu-8b94ab96_measuring-someones-productivity-by-their-activity-7468370692036526081-oB2k
- https://www.linkedin.com/posts/jia-rong-wu_introducing-devin-desktop-the-next-generation-activity-7467748343490949120-vWnc


Researched with Autumn (https://www.autumn.ai) from public sources. Updated 2026-11-04. Profile: https://www.autumn.ai/resources/cognition/people/jia-rong-wu · Everyone at Cognition: https://www.autumn.ai/resources/cognition/people