Leadingproduct, design and engineering teamsAnalyzing market and users unmetneedsArticulatingvalue propositionsBuilding productroadmaps and MVPsDesigning user experiences and building workingprototypesStory telling
Worked with
Big corporations , small startups and as a freelancerB2B, B2C and B2B2C products communication, advertising, health-care and other industries
Skilled in
Python Javascript SQLHTML CSS Neural networksLarge Language ModelsData analysis
Assuring that images uploaded by patients as part of a remote clinical consultation are of good quality to optimize clinicians time and consultation quality
An innovative solution to automate the process of connecting patients to clinicians, by creating a virtual waiting room and a predictive dialer for video and audio consultation
A truly killer feature that became key market differentiator and the main selling point of the product
The unmet need
Remote health service providers use video and audio to provide clinical consultations remotely
The prevailing practice is, that service requests flowing into the system are queued for a clinician
Clinicians pick a patient from the top of the queue and initiate the call
Which means that clinicians have to deal with unanswered calls, wrong numbers, patients not ready to take the call, etc
As a result, much of the clinician's time is wasted doing non clinical work, which leads to longer waiting times and overall inefficiency
The value proposition
A system that calls queued patients and puts them in a "virtual waiting room", enabling the clinician to click "Next" to join the call and immediately start the consultation
The KPI
Increase the ratio of clinicians "consultation time" to "connection time" versus a benchmark
Reducing manual resources needed to operate calls versus a benchmark
Keeping average patient waiting time in the virtual waiting room less than a target value
Increasing the system's throughput while reducing the operational resources
The main requirements
Should be workable 'out of the box' with predefined best practices configuration scenarios
Enable non-technical users to fine tune the basic scenarios to accommodated different business and operational use cases
Enable support / professional-services teams implement low level customization to the basic logic, without installing a new product version
The system logic
The overall predictive dialer logic
Predictive dialer logic (To enlarge, right-click the image and select "Open image in new tab")
The waiting room logic
Waiting room logic (To enlarge, right-click the image and select "Open image in new tab")
Trigger and fetch methods
The 'trigger' and 'fetch' methods are two functions at the heart of the virtual waiting room logic
They determine the logic by which appointments are selected from the queue to start the calling process, and how they will be fetched
from the virtual waiting room once a clinician is available
A predefined list of trigger and fetch methods are delivered with the product
Additional methods can be configured and quickly added per specific use case
User experience
Administrator
Service administrators can define the dialer settings, including the trigger and fetch methods, and the waiting room settings.
Configurations are made per service and can differ from service to service
Predictive dialer settings panel
Clinician
All the clinician has to do is click "Next" to start the conversation with the patient
Clinician 'next appointment' screenClinician video conversation screen
Patient
The patient receives a call from the system, and is asked to join the waiting room. Once the clinician is available, the patient is
automatically connected to the clinician
Patient virtual waiting room screen
Patient video consultation screen
The impact
The virtual waiting room and the predictive dialer became the core capability of the product, and the main market differentiator.
It's unique capabilities and fit for purpose, enabled organization to increase the number of patients served per time unit
by 50% to 200% in some markets, reduce the operational
costs by 20% - 40%
Image quality assurance
A computer-vision based system, that checks the quality of uploaded images in a remote clinical consultation, to optimize clinician / patient time
and the quality of the consultation
The unmet need
In remote clinical consultation, patients are often required to upload images prior to the consultation
The data has shown that 25% to 30% of uploaded images are of low quality
That leads to sub optimal use of clinician time and hinders consultation quality
The value proposition
Enhancing the process of image upload, adding a preemptive image quality check, prompting the user
to try again if the quality of the upload candidate is low
The KPI
Reducing the percentage of consultations cancelled with the reason "Bad images"
Benchmark: 30% cancellations as "Bad images"
Target: 10% or less
Business goals alignment: optimizing clinical resources efficiency and consultation quality
The acceptance criteria
85% or higher accuracy
Performance: under 1 second
Processed entirely on the edge device, hence eliminating network latency and servers load
The research
Training and testing data
A sample of 1000 uploaded images had been extracted
Of which 672 images had been tagged for quality as "Good" (24%), "Okay" (44%) and "Bad" (27%)
Model selection
A number of models were tested, including: BRISQU, NIMA and several pretrained neural networks that were additionally trained on the tagged images.
Eventually OpenCV was selected as it gave the best results and passed the acceptance criteria
Image assessment workflow
The workflow of assessing the image quality is comprised of 3 steps: processing the raw image, scoring the processed
image and then classifying the score into "Good" and "Bad" categories
Image assessment workflow
The prototype
To verify the assumptions and fine tune the model and user experience, the full concept was implemented as a Progressive
Web App prototype.
The image quality check had been implemented into the image upload process in the native mobile apps and the Web app. Feature
setting were added to the Admin Console app, to enable administrators fine tune the behavior to specific use cases / services
Image upload process with quality check
Image quality check integrated into the image upload process
Patient UX
Low quality image alert on patient app
Administrator settings
Administrators can:
- Toggle the image quality check on or off per service
- Select the scoring model (for additional future models to suit differen use cases)
- Set the max number of failed upload attempts (after which the image will be uploaded as is)
Image quality admin settings
The impact
A MVP was implemented in the Web app and the native mobile apps, exceeding the minimum required KPIs and a good level of user acceptance
Appointment rescheduling agent
A customizable agent that scours calendars and shift rosters, identifies under booked or over booked slots, and sends rescheduling - forward or backward in time - offers
to customers via email, SMS, phone or push notification, to fill the gaps and balance the load
The unmet need
Appointment based services ("scheduled services") face challenges in ensuring efficiency in a dynamic environment
For example, a scheduled appointment is canceled by the customer, which creates an empty slot or,
a service provider cancels an availability slot where appointment have already been scheduled
Such naturally occurring events make keeping an efficient and 'fair' system a challenge.
Usually requiring costly human intervention in order to review the changes and try to balance capacity vs demand while
maintaining the service level agreements (SLAs)
The value proposition
A system that reviews the calendars, detects unbalanced capacity / demand sections, and issues reschedule offerings to
customers, whether moving the appointment forward or backward in time
Example use case
A patient cancels a future scheduled MRI scan appointment
The system looks for appointments scheduled for a later time as candidates for filling the vacant slot
The system sends an email/SMS/voice message to the customers of those appointments,
suggesting to reschedule the appointment for the vacant time slot
When an offer is accepted by one of the candidates, the appointment is rescheduled and the vacant slot is filled
The KPI
Reduce the 'over / under capacity' (where service resources are idle or overbooked) time for a given time period
Reducing manual resources needed to operate the schedules
The main requirements
Should be workable 'out of the box' with predefined best practices configuration scenarios
Enable non-technical users to fine tune the basic scenarios to accommodated different business and operational use cases
Enable support / professional-services teams implement low level customization to the basic logic, without installing a new product version
The system logic
The system consists of two main processes: backward scheduling - which is the action of moving appointments backward in time, and
forward scheduling - which is the action of moving appointments forward in time
Backward rescheduling logic
Backward rescheduling logic (To enlarge, right-click the image and select "Open image in new tab")
Forward rescheduling logic
Forward rescheduling logic (To enlarge, right-click the image and select "Open image in new tab")
The user experience
Customer
Customers receive rescheduling offers via Email, SMS, phone or push notification based on the service settings
and can approve or reject the offer
Rescheduling agent communication channels Rescheduling agent customer UX flow
Administrator
Administrators can decide if to activate or not the rescheduling agent per service, fine tune its settings
and logic, choose the communication channels and the templates to be used and customize the messages. In addition,
the administrator can view the rescheduling agent's performance and efficiency reports
Rescheduling agent general settingsRescheduling agent email settingsRescheduling agent approval page text editing
The impact
The feature was released to selected test services, and showed a significant improvement in the efficiency of the appointment scheduling process.
Due to the complexity of the logic that needs to be implemented per use case, the test run raised the need for a simulation tool to test the results
of such a system before deploying it to production.
Virtual queue simulation
A virtual queue simulation tool that enables to test different configuration scenarios and their impact on the system's performance.
The tool became the go-to practice when introducing new configurations as part of providing professional services, and an addition to the
value proposition in the sales process
The unmet need
As a platform designed to automate high volume remote and face-to-face clinical encounters in many different environments and settings,
the way it is configured and customized is critical to the success of the deployment
As it grew in features and complexity, it became impossible to predict how different configurations will affect parameters
like waiting times, service times, queue lengths, clinician utilization etc. And in turn, how would that affect SLAs and patient clinical safety
The common practice was to deploy changes to production and pray that nothing goes wrong
Hence the need has arised for a reliable way to test different configuration scenarios before they were
implemented to production
The value proposition
A virtual environment that simulates the real world, and allows to test different configuration scenarios
and their impact on the system's performance.
The KPI
Simulated results are within reasonable accuracy with actual results
Adaption of the practice of simulation by customers and customer facing teams
The MVP
As an MVP the simulator was designed to be a stand alone tool for internal use
No graphical user interface required
Outputs can be exported and communicated with stakeholder
The simulator components
The simulator components (To enlarge, right-click the image and select "Open image in new tab")
The implementations
The simulator is implemented as a stand alone Python script that can be run from the command line. The outputs were both numerical and graphical
The simulation input
The simulation input example (To enlarge, right-click the image and select "Open image in new tab")
The simulation output
Some examples of the simulation graphical outputs
The simulation output example (To enlarge, right-click the image and select "Open image in new tab")The simulation output example - cont.(To enlarge, right-click the image and select "Open image in new tab")
The impact
The simulator proved to be highly accurate in predicting the system performance based on various configuration scenarios, an achievement
that was not possible before. It became the go-to practice when introducing new configurations internally and while communicating changes with customers.
Patient 360° view
A framework integrated into the core product that offers a simple, reusable, and scalable way to integrate and display data from multiple
sources in a single, customizable view
The unmet need
Telemedicine providers require - for both clinical needs and regulatory compliance - to view data about a random patient
from different sources
This data could be demographic - e.g age, gender, address; clinical - e.g existing conditions, treatments, prescriptions;
or other such as eligibility, insurance, child safety etc.
Usually there are multiple systems a clinician or administrator needs to interface in order to view the data,
which is inefficient. In the better case, an expensive bespoke integration is done which is hard to maintain and upgrade
Our product, bening the primary system used by remote and out-of-hours health care providers, clearly needed
a simple, reusable and scalable way to integrate and display data from multiple systems
in a single view, which is easy to implement, maintain and reuse
The value proposition
A framework, built into the core product, that enables a structured way to integrate and display data from multiple systems,
which is easy to implement, customize to specific customer's needs, maintain and reuse
The KPI
The KPI from a B2B product point of view is the amount of customers opting for an integrated
solution vs using multiple systems
The main requirements
Should define a customizable structure for displaying summarized data and detailed data
Should define a structured 'plug in' mechanism to add data resources
Should be customizable, within defined boundaries, to specific requirements
Conceptual architecture
The system consists of reusable data sources plugins, a customizable data processing engine and a templates based display front end
Patient 360°, conceptual architecture (To enlarge, right-click the image and select "Open image in new tab")
Display framework
The display framework consists of a designated place for the summarized data and a place and structure for the detailed data,
controlled by customizable templates
Patient data display framework (To enlarge, right-click the image and select "Open image in new tab")Summary data display exampleDetailed data display example
Administration
Administrators can select suitable display template per service
Service template selection
The impact
The framework prooved to be highly successful in enabling the integration of multiple data sources into a single view, and adopted by multiple customers
In one case, when a customer suffered a prolonged outage of its clinical data system, our teams managed to integrate the sources into our system in a few
days utilizing the framework, enabling the business continuity of the customer
Shift planning
An integrated shift planning tool that leverages data to enable better planning quality, simplify the process, provide insight
that is otherwise not available to the planners and enable additional automation of other processes
The unmet need
Clinical practitioners are typically the most in-demand, in-shortage and costly resource for health service providers
Hence, carefully planning their shifts is critical for maintaining an efficient and viable service
In our target market, shift planning was done with various tools from white board and markers to spreadsheets to various specialized software
All of the aforementioned tools demanded onerous effort, and depended heavily on human knowledge of the day to day operations in order to be efficient
The value proposition
Having large amounts of data on the actual operations of the clinical services, we saw an opportunity to provide an integrated shift-planning tool,
leveraging data to enable better planning quality, simplify the process and provide insight that is otherwise not available to the planners
The KPI
Adoption level of the shift planning tool by services
The implementation
Clinicians' shifts are entered via a graphical user interface per service
Based on the entered shifts, the system calculates the aggregated capacity derived from the individual characteristics of each assigned clinician
The system presents the total capacity vs the expected demand based analyzing historical data, enabling to balance the two
In addition, for every shift, the system calculates the predicted waiting times for the given planned capacity and predicted demand,
enabling insight on SLA compliance
Clinician shifts as entered per service (To enlarge, right-click the image and select "Open image in new tab")Viewing clinicians data (To enlarge, right-click the image and select "Open image in new tab")Analysing demand vs capacity per shift (To enlarge, right-click the image and select "Open image in new tab")Presenting the waiting times based on the given configuration(To enlarge, right-click the image and select "Open image in new tab")Calculating the capacity per hour (To enlarge, right-click the image and select "Open image in new tab")Calculating the waiting times (To enlarge, right-click the image and select "Open image in new tab")
The impact
The tool was adapted with good user satisfaction in some services, but faced challenges in use cases where the organization was heavily invested in a third
party specialized solution which offered more features. Additional development was required to successfully compete with the specialized solution
Owing to his unique and extensive skill-set, Erze is Instrumental in the conceptualization
and design behind all advanced and innovative features. In addition, he also took a dominant role
and became instrumental in providing crucial knowledge expertise for the PS/account mgmt/support teams.
Within the product team, Erez is the main simplifier, problem solver, rational thinker and
counter balancer. I can confidently say that “No Erez, less of a chance to nail a MVP”
On a personal level, I enjoy working with him and often look forward to our conversations
to share thoughts and hear his feedback, both on work and non work related issues
In summary
Erez is instrumental for me to be able to deliver on our product vision and
objectives, as well as for the current, and future, success of the company. I think he is a
mission critical asset to the company both in terms of his personality and his extensive
and unique skill-mix, experience and capabilities.
Dr. Eyal Eaglehearted, Ummanu
I had the pleasure of working closely with Erez for several years on a complex telehealth digital product,
where he led product management and I handled product design.
Throughout our collaboration I was constantly impressed by his methodical approach, deep understanding of user needs,
and ability to navigate complex challenges with clarity and precision.
Erez is not only a great listener—always open to feedback and different perspectives—but also a strong strategic
thinker who brings brilliant ideas to the table in an effective and actionable way. His ability to balance user
experience with business goals made our collaboration seamless and highly productive.
Beyond his skills, he’s simply a pleasure to work with—professional, thoughtful, and a true team player.
Any company would be lucky to have him on board!
Adam Shriki, Product designer
I had the pleasure of working alongside Erez for close to 6 years, when we were part of the management team at Netex.
Erez always had a unique perspective and approach which had a tremendous effect on the development and product teams.
His ability to take ideas and manifest them into fully operational products was remarkable.
He was involved hands-on with every aspect of the product journey, understanding the clients' needs, highlighting potential issues, and providing solutions.
A true professional at his craft. I hope we get to work together again in the future
Eitan Gissin, Sade Biotech
I had the privilege of working with Erez at Ummanu for almost six years, collaborating on the development of numerous innovative and valuable
features for our telemedicine platform.
This deep understanding of client needs and ability to translate them into actionable product stories were exceptional.
His structured, insightful, and highly professional approach to storytelling helped me and the entire R&D team fully grasp the objectives
behind each feature. This clarity significantly streamlined the development process and enabled us to deliver effective, well-integrated
solutions with confidence.
Collaborating with Erez was a fantastic experience. His expertise, strategic mindset, and ability to deliver impactful solutions make him
an invaluable asset to any organization.
Arsenii Krivoshei, Ummanu
If you are looking for a true creative professional, then Erez Dori is your man. He meticulously went above and beyond my expectations
in an efficient and timely manner, when previously others told me the scope of work would be too complex.
I was equally impressed with his patience and customer service
Dragan Vistica, Entrepreneur
Dear Erez,
The time has come to write you this letter as a sign of appreciation for your work at "Teleconsulting".
From your first day with our company, you have shown initiative, diligence and responsibility which have contributed significantly to the company's success.
You have demonstrated complete control over all aspects of our work, starting from cultivating reliable relationships with our customers, through independent
development of complex solutions, and ending with training our teams in all the aspects of the fruits of your labor.
In the last year, you have contributed another layer as the manager of the consulting group, in the integration of a new system and its successful implementation
within the company.
Your unique personal qualities, like your excellent professional skills, place you at the forefront of professionals in your field.
I am confident that you will succeed in whatever you set your mind to. Please consider this letter as a warm recommendation for whom it may concern
