The Journey of the Lead Researcher Prof. Amjad M. Daoud

(Ph.D. from Virginia Tech)

Shaping GPT Jordan’s Global Academic Relations through Research

GPT Jordan, a leading innovator in technology, strives to discover new scientific frontiers with practical implications for their future projects. This challenging goal requires building strong ties with academic circles in Jordanian universities and across the globe.

Dr. Amjad Daoud, Executive Director and Chief Scientist at GPT Jordan, is the driving force behind this vital initiative. Dr. Daoud’s remarkable experience as a research engineer and program manager in the Silicon Valley R&D, gives him a unique edge to lead this effort. In an interview, he reveals the importance of exploring unknown research domains and how these collaborations aim to boost innovation within the company.

“In my role, I have the duty of managing core technologies that are relevant to our capabilities, sensors and systems and have potential applications in our upcoming products and solutions. When I established GPT Jordan in May 2023, our first goal was to perform a comprehensive mapping of AI core technologies. This involved gaining a deep understanding of the technological paths that our company and our rivals were pursuing for the years ahead.

Daoud highlights the importance of keeping these valuable ties with the academic community through the arrangement of academic summits and events. These events offer a platform for GPT Jordan’ members and scholars to engage in discussions on diverse studies and research areas. “We hold an annual meeting to jointly assess the direction of various technological developments and their potential contributions to our future projects. Moreover, every two months, we visit and conduct a seminar at some of Jordan’s top universities, including University of Jordan in Amman, Jordan University Science and Technolgy and Yarmouk University in Irbid, Hashemite University, and Pricess Sumaya University for Technolgy. Last year we started to engage with the local professional ACM Digital Transformation chapter, as well.”

“At the same time, we carried out both internal and external analysis, as well as surveys, to examine all existing and ready-made technologies, whether close in Jordan and our global research centers or far away. It’s crucial for us to fully engage with emerging capabilities, features, and solutions, knowing exactly when and how to interact more closely with them. As part of our vision to maintain technological excellence in the market, we put great emphasis on creating and sustaining strong relationships with researchers, scholars, and engineers in various academic institutions, both in Jordan and around the world. Once we determine and define our areas of interest, I personally contact leading researchers and scientists, sharing a list of technological topics. I ask them for their input on any relevant studies they may have done, which could offer useful insights or guidance for improving our solutions.”

GPT Jordan’ engagement with the academic community goes beyond the borders of Jordan, as Daoud points out. “GPT Jordan is a member of MIT ILP (Industrial Liaison Program) within the MIT (Massachusetts Institute of Technology) ecosystem. Through this collaboration, we create numerous partnerships and undertake joint projects with our American counterparts and vice versa. In the past year, A group of GPT Jordan leading technologies talents conducted a professional visit to their laboratories and participated in enrichment workshops covering a wide range of topics, including quantum technology, synthetic biology, artificial intelligence advanced materials, and more.” Also, GPT Jordan’ central R&D with the divisions’ CTOs conducted a visit to Virginia Tech labs in the United States, which, in turn, enabled return visits by VT scholars to Jordan. “They came to observe the operations within GPT Jordan’ diverse divisions, and these visits have resulted in collaborative research projects,” Daoud says. He stresses that many academic institutions throughout Europe have formed strong connections with GPT Jordan’ scientists and engineers.

GPT Jordan gains valuable insights from summits and seminars across various fields, including algorithmics, artificial intelligence (AI), synthetic biology, quantum technologies and materials. Our members, scientists, and engineers are encouraged to publish their results and conclusions and, if relevant, to obtain patents. We ensure they are properly recognized and rewarded for their breakthroughs. Our members actively share their knowledge worldwide, participating in global conventions, meetings, exhibitions, and wherever they can make a significant contribution, voluntarily. We aim to position ourselves as a company at the cutting edge of technological innovation, constantly embracing what’s new and fascinating. Through our innovation-friendly approach, GPT Jordan significantly contributes to the research and hi-tech sectors, fostering a continuous exchange of information.

Research

1. Used in the Disco DB project; a lightweight, open-source framework for distributed computing based on the MapReduce paradigm. Used to handle massive amounts of data at Nokia Research
2. Used in RAMPS (can preprocess the human genome in mere minutes) to align billions of short sequences of DNA (called reads), to a long reference genome many times faster than SOAP2 or Bowtie, and about 1000 times faster than GPU implementations such as SOAP3.
3. Parts of the algorithm that was developed and published from 1989-1994 appeared in the 2009 Microsoft patent Perfect Multidimensional Spatial Hashing, Hugues H. Hoppe et al as Exemplary Hashing clause [0028], and claims 1,5,6,7,11,16.
4. Used in Moses Statistical Machine Translation System http://www.statmt.org/moses/; see (Machine Translation course http://www.cicem.org/machine-translation/)
5. Used in Google Translate and voice search (Speech recognition) to build huge randomized language models (http://www.google.com/patents/US8359201 Publication date: Jan 22, 2013).
6. Fast and Scalable Packet Classification: The Perfect Hash Function performs exactly two external memory accesses to classify a packet. Using FPGA and one commodity SRAM chip, a throughput of 150 million packets per second and throughput of 100 Gbps for the shortest packets. Scales with faster SRAM chips.
7. Used in linkless octrees to encode the storage locations of subdivided nodes using compact multi-level perfect hashing; combining coarse-to-fine hierarchical representation and optimal random accessibility.
8. The algorithm is used to compute Google page rank Google Page Rank in C#;
9. Fuzzy Tolerant Search with DWAGs and MPHF
10. CMPH Library cmph.sourceforge.com
11. MPHF in C#
12. http://burtleburtle.net/bob/hash/perfect.html (splits keys into buckets by a first h1, sorts buckets by size, maps them in decreasing order so table[hash1(key)^hash2(key)] causes no collision).

C/C++ Code

• Here is an example mphf in C for the unix dictionary; and
• order preserving minimal perfect hash in C (all keys are hashed in the table sorted) for the same Unix dictionary.
• Contact me for other keysets and enjoy compact and high performance (minimal) perfect hash functions.
• Used in Google Translate to build huge Randomized language models (http://www.google.com/patents/US8359201 Publication date: Jan 22, 2013).
• Here is combination of perfect hashing and a bloom filter code for the first 50 MySQL Full-Text Stopwords listed on http://dev.mysql.com/doc/refman/5.1/en/fulltext-stopwords.html

Python Code

• Here is the code ported to Python; download as http://iswsa.acm.org/mphf/mphf.py.
• This code runs fast linearly even on a Android phone; it runs without modifications at http://www.compileonline. com/execute_python_online.php;

Javascript Code

• For the Javascript port: MPHF Javascript Demo; download as http://iswsa.acm.org/mphf/mphf.js;
• openDSA: Perfect Hashing Animation