Keyvan M. Sadeghi

AI/Software Engineer


Location: Toronto/Canada, Status: Work permit, Iranian national

I'm co-founder and CTO at Assister.Ai . My passion is Artificial Intelligence (AI); more specifically, materialisation of thinking machines ( AGI ). Previously, I mainly served in the capacity of technical lead and been responsible for replacing legacy systems/architectures with cutting (often bleeding) edge ones. I can apply Machine Learning at scale and elevate businesses via creating added value by adopting AI. I have a track record of being effective providing the R&D function in an agile setting, where novel ideas are transformed to tangible results and delivered in a timely fashion. I occasionally blog at Medium and .


University of Southampton, UK

MSc in Artificial Intelligence with Distinction, 2012


Machine Learning

Applying ML to consumer-facing products.

Software Architecture

Over a decade of first-hand experience on adopting cutting edge software paradigms.

Open Source

Releasing most of the foundations of my work as OSS. Contributions to others.


  • Python
  • JavaScript
  • Big Data
  • Graph Database
  • Containers
  • Cloud


Assister Technology LTD

Co-founder, Technology Lead


I moved to Canada as a recipient of Startup Visa , sponsored by NEXT CANADA , a highly reputable technology incubator in Toronto. We were participants of NextAI , their AI accelerator program, and received 30K CAD in investments.

We are building a Private, Open and General Assistant platform at Assister.Ai, check out the RFC and the progress at GitHub.

Mahan Airlines

Co-lead, AI Team


I joined Mahan as a Software Architect and helped with the migration of fleet assignment/planning legacy software.

I later started the Mahan AI team with Hamidreza Mahyar , we were responsible for the ROI and management of the team. The team expanded from 2 to 10 in the span of a year and one of the projects that we started ultimately automates fleet assignment at Mahan using Machine Learning.

MAPS Startup Studio

Technical Director


I helped startups adopt the Agile methodology while they reach POC and MVP stages by providing hands-on technical support, effectively being a member of their team until they graduate.

I joined Miras Technologies (4 US patents in Big Data) in my first year at MAPS. Helped with a product pivot right after the founder left. One of the top achievements was creation of a real-time predictive maintenance system on top of Apache Spark for analysing turbines of a power station.

In my second year, I moved to Satratech (an E-Health startup) and helped with revamping the backend of Shafajoo , a platform for matching and reservation for doctors and patients. Developed a Neo4j graph model for E-Health and created the koa-neo4j framework in the process.

My job involved regular meetings with MAPS' board, updating them on the current state of the 5 startups working there, filling the vocabulary gap between technical and business people.

Azad University of Isfahan /
Isfahan science and technology town

Lecturer / Startup Co-Founder


Teaching Foundations of AI to bachelor students.

Co-founded the "Cloud Computing Group" startup in ISTT university accelerator where we tried to offer IaaS and PaaS products to Iranian companies. This startup eventually failed since Iranian companies would not buy into cloud concepts at the time.

OpenCog ITF project /
Novamente LLC

Project Associate / Freelance AI Developer


I joined OpenCog in the ITF project at their Hong Kong lab right after graduation. Participated in design and development of many sub-components of OpenCog.

I continued working with Ben Goertzel on OpenCog as a contractor of Ben's company when I moved back to my home country. This collaboration resulted in four publications.



First Author


Uncertain Interval Algebra via fuzzy/probabilistic modeling

IEEE Symposium on Computational
Intelligence for Human-like Intelligence

Team Paper


The CogPrime Architecture for Embodied Artificial General Intelligence


Team Paper


Pattern Mining for General Intelligence: The FISHGRAM Algorithm for Frequent and Interesting Subhypergraph Mining


Team Paper


Syntax-Semantic Mapping for General Intelligence: Language Comprehension as Hypergraph Homomorphism, Language Generation as Constraint Satisfaction

Last updated May 2020 — [email protected] — (+1) - 416-859-8984