App & Software Development for Machine Learning Startup

Introduce your business and what you do there.

We are an Israeli startup working on motion recognition technology. We use machine- and deep-learning components to analyze people’s speech and detect what emotions they feel during a particular period. Our products, solutions, and services are based on this idea. I led our Ukrainian office for a time, and am currently a senior manager of the organization.

What challenge were you trying to address with INOXOFT?

We work with them to develop products and services based on our main idea and technology. We’ve developed machine- and deep-learning technology, and needed to develop internal tools and external solutions.

What was the scope of their involvement?

INOXOFT has developed our cloud solution using Microsoft Azure and ASP.NET, and frontend using AngularJS. They also developed Android and iOS applications for different products and services we offer.

We needed to make our machine learning component more efficient. We needed a tool that would allow us to feed different types of data to the machine learning engine, making it more accurate and faster.

We had a team that analyzed voice recordings and determined which emotions were being experienced at different times. INOXOFT has developed an internal software tool that allows us to upload voice recordings and track the emotions displayed in them. Our own employees use the tool. It was developed as a cloud solution with an Azure backend.

We needed to promote our offerings during an event and had the idea of creating an Android application. The app tracks phone conversations, calculates trends, and provides statistics such as how happy the user was during a call.

They've also developed an Android and iOS cloud application for hospitals. Prior to attending an appointment, patients can log into the application and answer some questions through text and voice recording. Based on that, we can give doctors predictions and analyses about what was happening with the patient.

What is the team composition?

We started by using a couple of resources from their team. We use them as needed. In total, we’ve used 10–12 developers, QA testers, and managers from their team.

I work directly with one or two managers from INOXOFT, depending on how much ongoing work we have. They led the teams from their side.

How did you come to work with INOXOFT?

One of our Ukrainian employees recommended the team after we asked for referrals. I got in touch with INOXOFT and discussed what they were offering. I decided to have a trial, using one or two developers from their team to create a small portion of an internal tool. I was very happy with their quality, delivery, and communication, so we started to expand the team and do more things.

I evaluated 10–12 different vendors in total, evaluating their teams and costs. I had multiple interviews, and INOXOFT gave us a very good cost/value proposal.

How much have you invested with them?

We have spent around $450,000–$500,000 on their services. We’ve had a large team for around three years.

What is the status of this engagement?

We started working with INOXOFT around three years ago.

What evidence can you share that demonstrates the impact of the engagement?

The internal tools we’ve developed with them work well. We have significantly improved the quality of the machine-learning engine. It also gives us a way of tracking employees and improving their performance. The UX is pretty good, and the application works smoothly. We and our employees are happy with the results. This confirms that INOXOFT performed well.

How did INOXOFT perform from a project management standpoint?

The communication is quite clear, and the managers I’ve worked with are responsive. I can communicate with someone whenever I need something urgently need or want a report.

I've visited them on a couple of occasions to ensure that we're in sync regarding our roadmap. This has worked very well.

Their managers do a good job and go the extra mile. I didn't expect them to work on the product itself as much, and provide some level of business analysis. They give us feedback and ideas for improving the product. They've persuaded me to choose a different approach on a couple of occasions. They don't just do what I say but also bring their ideas to the table.

We haven’t had significant delays. We've deviated from the roadmap a little, but they work hard to deliver.

What did you find most impressive about them?

Their management, responsiveness, communication, commitment, and work quality are all very good. They work hard to ensure that we meet our deadlines.

Internal employees evaluate the work, and we perform code reviews from time to time. They don't switch from senior to junior resources like many in the industry do. The quality stays at the same level. They're honest and transparent, which I like a lot.