I’m a Research Scientist with a strong technical background in HW/SW co-design and Edge AI.
In summary I’m:
- A Research Scientist at Rain
- The creator of FedSim Simulator (https://fedsim.varnio.com)
- The first inventor of an international patent on Transfer Learning
- A reviewer for top ML conferences like ECCV, CVPR, ICCV
and I have:
- 4+ years experience as Research Scientist (industrial)
- 1+ years experience as Data Scientist and Research Assistant (academic & industrial)
- 2 years experience as FPGA Engineer/RTL Designer (industrial)
Research & Industry
Research Scientist, Rain
August 2023–present, Montreal, Canada
- Research and development of novel machine learning algorithms for edge devices
Consultant, ML Engineer, Varnio Tech.
Jul 2022–July 2023, Montreal, Canada
- Consult with clients to develop and implement distributed, edge, and federated machine learning solutions
- Provide expertise in machine learning model optimization and deployment
- Communicate technical concepts to non-technical stakeholders
- Collaborate with cross-functional teams to ensure successful implementation of machine learning solutions
- Stay up-to-date with industry trends in machine learning and best practices
Research Scientist, Imagia
May 2018–March 2022, Montreal, Canada
- Research on Federated Learning optimization led to SoTA performance via drift elimination
- Research on Transfer Learning led to a filed patent
- Research on multiple Meta Learning and Few-shot Learning projects
- Research on Multi-hypothesis Transfer Learning and out of distribution generalization
- Collaborated with R&D team in designing an AI library for Imagia research
- Collaborated with IT in porting Polyaxon on a cluster of NVIDIA DGX systems
May 2017–May 2018, Halifax, Canada
- Research on predicting human behaviour from fMRI data
- Developing a CNN framework for detecting corrosion in aircrafts using D-Sight technology (DAIS)
- Optimizing calculation of minimum distance to shore from AIS-GIS streaming data using CUDA and OpenMP
- Research on sparsity, activation functions and normalization
Mar 2020–Jun 2020, Halifax, Canada
- Clinical data integration and visualization
- Investigating post-operative cognitive dysfunction in elderly patients
- Analyzing surgical time series data (anesthesia depth, patients’ vitals, …)
Jun 2013–Jun 2014, Tehran, Iran
- Design & Imp. of TDM switches on FPGAs supporting up to 16k x 16k channels (in VHDL)
- Multi-channel I2C master controller supporting 16 modules with error checking & correction
- SPI & USART Peripheral interfaces
- Embedded Processors, RTOS
- Focus on speed optimization on Altera Cyclone series
Jul 2012–Jun 2013, Isfahan, Iran
- Design & Imp. of lightweight AES modules used in STM4 lines
- Multi-channel I2C master controller supporting 16 modules with error checking & correction.
- SPI & USART Peripheral interfaces.
- Focus on area optimization on Xilinx Virtex 4, 6 series
- Co-instructor, ML for Big Data, CSCI-6515, Dalhousie University, Fall 2020
- Teacher Assistant, ML for Big Data, CSCI-6515, Dalhousie University, Fall 2018
- Teacher Assistant, Digital Circuits, ECED-2200, Dalhousie University, Winter 2016
- Teacher Assistant, System Analysis, ECED-3401, Dalhousie University, Fall 2016
- Instructor, Computer Architecture, Chehelsotoon Inst. for Higher Edu, Fall 2015
- Instructor, System Programming, Chehelsotoon Inst. for Higher Edu, Fall 2015
- Teacher Assistant, Java Programming, University of Guilan, Winter 2009
- Teacher Assistant, Algorithms, University of Guilan, Winter 2010
- Ph.D., Computer Science. Dalhousie University. 2017–2023, CGPA: 4.19
- M.Sc., Computer Architecture. University of Isfahan. 2012–2015, CGPA: 4.02
- B.Sc., Guilan University. 2008–2012.
- Programming languages: Python, Java, C/C++, Bash
- Deep learning frameworks: PyTorch, Keras, Tensorflow
- CI/CD platform: Github Actions
- MLOps, automation & AI scaling systems: Polyaxon, MLflow
- Target specific on-device training, Quantization & Compression
- Machine learning libraries: Pandas, Scikit-learn, Numpy, Scipy
- Digital Circuit Design and FPGA engineering: VHDL, Verilog
- Markup languages: \LaTeX, Markdown, RestructuredText, Mermaid
- Project Management: Agile, Scrum, Kanban, Jira, YouTrack
Varno, Farshid, Marzie Saghayi, Laya Rafiee, Sharut Gupta, Stan Matwin, and Mohammad Havaei. “Minimizing Client Drift in Federated Learning via Adaptive Bias Estimation.”
European Conference on Computer Vision. – ECCV (2022).
Varno, Farshid, Lucas May Petry, Lisa Di Jorio, and Stan Matwin. “Learn Faster and Forget Slower via Fast and Stable Task Adaptation.” arXiv preprint arXiv:2007.01388 (2020).
Varno, Farshid, Behrouz Haji Soleimani, Marzie Saghayi, Lisa Di Jorio, and Stan Matwin. “Efficient neural task adaptation by maximum entropy initialization.” arXiv preprint arXiv:1905.10698 (2019).
Jiang, Xiang, Mohammad Havaei, Farshid Varno, Gabriel Chartrand, Nicolas Chapados, and Stan Matwin. “Learning to learn with conditional class dependencies.” In international conference on learning representations. – ICLR (2018).
Saghayi, Marzie, Jonathan Greenberg, Christopher O’Grady, Farshid Varno, Muhammad Ali Hashmi, Bethany Bracken, Stan Matwin, Sara W. Lazar, and Javeria Ali Hashmi. “Brain network topology predicts participant adherence to mental training programs.” Network Neuroscience 4, no. 3 (2020): 528-555.
Leadership & Volunter Work
- Vice-president of Public Relations, Toastmasters International, Dal Toastmasters, 2020.
- Experienced leading teams of 2-3 researchers during several projects.
- Mentored two masters students, now working as Senior Data Scientists in USA & Brazil.
- Reviewer at International Conference on Computer Vision (ICCV 2023, upcoming).
- Reviewer at Computer Vision and Pattern Recognition (CVPR 2023, 5 papers).
- Reviewer at European Conference on Computer Vision (ECCV 2022, 2 papers).
- Reviewer at the 2nd FedVision Workshop (FedVision 2023, 2 papers).
- Selected conference program committee member & volunteering experience:
- International Conference on Learning Representations (ICLR, Remote 2020)
- SIGKDD Conference on Knowledge Discovery and Data Mining (KDD, Halifax 2017).
- Confoo (Montreal, 2023).
Awards & Recognition
- Accelerate Award, 56k CAD, Mitacs, 2021-2022
- Scotia Scholar Award, 45k CAD, Research Nova Scotia, 2019-2021
- Best Graduate Student Research Award, Big Data Congress, Sep 2017
- Nova Scotia University Student Bursary, Government of Nova Scotia, 2020-2022
- FGS’s alloc. for outstanding status, 2k CAD, Dalhousie University, Aug 2017
- 1st Rank Student Recognition, University of Isfahan, Mar 2015