Stealth mode AI startup

• Designed an AI workflow to create a costum text guided 
segmentation model (VLM).
• Create two unique low computation vision language action (VLA) model
(one is going to be a publication)
• Distilled knowledge for LLM 
• Design and implement workflow for Agentic AI 
• Designed a multimodal DL/ML workflow for image and LiDar fusion for aerial imagery
•  Federative learning for collaborative robots, e.g. drones and autonomous driving vehicles

Principal AI engineer

Sanborn

•  Lead AI team, hands on projects and create innovative approaches which saved couple of
millions of dollars.
•  Leading the research on LLM, generative AI and traditional CNN
• Designed a AI workflow to create an HD-map with accuracy less than
20cm.
• Create a unique approach to georeference topological maps utilizing deep
learning/machine learning/computer vision and LLM. My team in Sanborn
won fourth, an honorable mention, place in AI for critical mineral assessment
competition organized by DARPA and USGS
• Designed a DL/ML workflow to extract information from aerial and
satellite imagery

VP of AI/ML

HERE Technologies

• Working in an autonomous driving team:
• Designed and implemented perception for mobile camera (dashcam).
• Designed and implemented the real-time perception for autonomous
robots 
• Lead the design and implementation of localization pipeline for
autonomous robots

Principal Data Scientist AI/ML

Stealth mode startup

• Leading the AI team and designed and implemented the pipeline for object
detection and tracking for indoor and outdoor environment
• Designed and implemented SLAM for indoor environment (ongoing project) 
• Designed and implemented a prototype of a new sensor for localization 
• Deep investigation to adopt new algorithm for localization 
• Designed and implemented the pipeline for object pose detection and reduce the cost of
hardware from 20k$ (Lidar/s) to less than 1k$ by using camera/s.
Note: No longer in the business.

Head of computer vision and sensor fusion

BYTON

• Filed five patents related to AI for AD • Lead the computer vision team • Invented an approach which makes it possible for vehicles with hardwares suitable for ADAS level 3 to upgrade to
level 4 of autonomy with minimum changes in vehicle network architecture.
• Invented a perception redundancy to increase safety
of the driver-less vehicles.

Tech lead/manager AI/AD

Tech lead Autonomous driving

• Designed a unique pipeline for parallel parking of AVs • Filed four patents in the area of sensor for autonomous driving (one), deep/machine learning(one) and ADAS/AD (two). • Developed and designed a unique HD map for driver-less
vehicles. • Lead the team to implement the special designed HD map including
the static perception required for HD map. • Implemented vehicle
localization and fused the required sensors. • Invented two new sensors for
AD cars. Co-invented a deep learning and machine learning algorithm.
Note: Byton was Defuncted.

Lead engineer for AD platform

DAQRI

Solid state Lidar: 
Developed and implemented FMCW LiDar (short range). Proposed an innovative (prospective patent) technique to overcome the problem of slow raising time for Liquid Crystal on Silicon (optical phase arrays).
Developed and implemented Convolutional neural network (deep learning) for cloud points.

Lead software engineer

Solid state LiDar
Developed and implemented Localization algorithm for AR device in Linux environment.
Compressed sensing: developed and implemented algorithm to recover original scene from sparse data (cloud points).
Note: Daqri was dissolved.

Sr Software engineer

Velodyne Lidar, Inc.

• Played a key role in performance enhancement and algorithm development for
VLP-32 products • Designed, implemented, and verified Long Shot (LS)
algorithm to extend detection range of LiDAR sensors beyond 200 meter • Modeled
and tested existing firmware detection algorithms • Considerably improved system
SNR through statistical and theoretical analysis • Studied object detection and
receiver operating characteristic (ROC) of the system • Successfully performed
troubleshooting and debugging of the sensor using data analysis techniques
• Applied modified machine learning algorithms and identified anomalies in the lasers
• Successfully modeled and tested two algorithms for suppressing interference in
LiDAR systems with dual laser firing using MIMO and uniarray antenna. In separate
study, optimized firing sequence of the lasers to minimize the interference.
• Formulate and document existing firmware calibration algorithm

Software Engineer

Cepand, Inc.

Sep 2014- May 2015: Embedded system for 1) IoT and 2) Kalman filters
Jan 2014- Sep 2014       : I was involved in two different projects: 1) Massive MIMO in collaboration with NCIC labs at UC, Irvine 2) Indoor positioning 

2013-2014: Location based algorithms for seismological and geological applications. 

Result: Two provisional patents are submitted to USPTO.

UC Irvine

Supervisor: Prof. Hamid Jafarkhani

Research title :Optimization of GPS receivers related algorithms. 

Result: 

a) A published conference paper in Asilomar 2013.

b) Two published journal papers in collaboration with TUT.

Visiting Researcher

• End-to-end AI/robotic for autonomous driving/robotic including HD-map, perception, localization, prediction and planning.
• 5+ years of experience in detection/estimation algorithms for AD/ADAS sensors (GPS,
LiDAR, and Ultrasonic) for self-driving vehicles with ten patent. 
• Supervised/unsupervised learning, Generative AI, Neural network, LLM, Deep learning, Support Vector Machines, Clustering, Principal component analysis, Recommender Systems
• 10+ years of experience in Digital Signal Processing, algorithm design and system
modeling in Communications and optical Systems
• Experience in design and implementation of machine learning and data science algorithms (supervised and unsupervised).
• Strong analytical, mathematical, and problem solving skills
• Solid knowledge of design and verification of Digital Filters, multirate and statistical signal processing. 
• Good knowledge of Spectrum sensing methods (applied to cognitive radios), indoor positioning techniques (DOA, AOA, ...)
• Programming skills (Python, C++, Matlab/Octave, R)
• Experienced in programming and working w/ TI DSP and Arduino.

I’m hiring! If you’re passionate about the AD industry and eager to join an amazing R&D team working on cutting-edge technology, DM me — we’d love to connect.

taplio

Becky Soltanian, PhD's Linkedin Analytics

Becky Soltanian, PhD

Performance ScoreBecky Soltanian, PhD's LinkedIn Performance

Becky Soltanian, PhD's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Becky Soltanian, PhD's Linkedin Analytics

Becky Soltanian, PhD

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Becky Soltanian, PhD's LinkedIn Performance

Becky Soltanian, PhD's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Becky Soltanian, PhD's Linkedin Analytics

Performance ScoreBecky Soltanian, PhD's LinkedIn Performance

Becky Soltanian, PhD's Best Posts (last 30 days)