INOD Innodata Inc

MANAGEMENT'S DISCUSSION AND ANALYSIS OF FINANCIAL

CONDITION AND RESULTS OF OPERATIONS

Disclosures in this Form 10-Q contain certain forward-looking statements, including without limitation, statements concerning our operations, economic performance, and financial condition. These forward-looking statements are made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. The words “project,” “head start,” "believe," "expect," “can,” “continue,” “could,” “intend,” “may,” “should,” “will,” "anticipate," "indicate," "point to," “forecast,” “predict,” “likely,” “goals,” “optimistic,” “foster,” “estimate,” “plan”, “potential”, or the negative thereof, and other similar expressions generally identify forward-looking statements, which speak only as of their dates.

These forward-looking statements are based largely on our current expectations and are subject to a number of risks and uncertainties, including without limitation, that contracts may be terminated by clients; projected or committed volumes of work may not materialize; the primarily at-will nature of contracts with our Digital Data Solutions clients and the ability of these clients to reduce, delay or cancel projects; continuing Digital Data Solutions segment revenue concentration in a limited number of clients; continuing Digital Data Solutions segment reliance on project-based work; inability to replace projects that are completed, canceled or reduced; our dependency on content providers in our Agility segment; difficulty in integrating and deriving synergies from acquisitions, joint venture and strategic investments; potential undiscovered liabilities of companies and businesses that we may acquire; potential impairment of the carrying value of goodwill and other acquired intangible assets of companies and businesses that we acquire; changes in our business or growth strategy; depressed market conditions; changes in external market factors; the ability and willingness of our clients and prospective clients to execute business plans which give rise to requirements for our services; changes in our business or growth strategy; the emergence of new or growing competitors; various other competitive and technological factors; and other risks and uncertainties indicated from time to time in our filings with the Securities and Exchange Commission.

Our actual results could differ materially from the results referred to in the forward-looking statements. In light of these risks and uncertainties, there can be no assurance that the results referred to in the forward-looking statements contained in this Form 10-Q will occur and you should not place undue reliance on these forward-looking statements. These forward-looking statements speak only as of the dates such statements are made.

We undertake no obligation to update or review any guidance or other forward-looking statements, whether as a result of new information, future developments or otherwise.

Business Overview

Innodata Inc. (NASDAQ: INOD) (including its subsidiaries, the “Company”, “Innodata”, “we” or “our”) is a global services and technology company. We combine human expertise with deep learning technologies to power leading information products and enterprise artificial intelligence (AI)/digital transformation.

The Company, founded in 1988 and headquartered in northern New Jersey, features a 3,500-strong global delivery and technology team spanning ten offices globally and a research and technology incubator, Innodata Labs, which focuses on applied machine learning and emerging artificial intelligence.

Our core services are (i) data acquisition, transformation, and enrichment at scale; (ii) digital operations management and analytics; and (iii) content applications. We report our core business as the Digital Data Solutions (DDS) segment.

We also have venture businesses that leverage our core capabilities to provide specific industry solutions. Our Synodex venture business delivers a software-as-a-service (SaaS) platform and managed services for digital transformation of medical data. Our Agility PR Solutions (Agility) venture business delivers a SaaS platform and managed service for delivering news, information, and content to targeted journalists and influencers as well as monitoring and analyzing coverage across traditional and social media sources. Each venture business is reported as a separate segment.

Our Core Business (Digital Data Solutions (DDS) Segment)

We specialize in combining deep neural networks and human expertise in multiple domains (including health, science, and law) to make “unstructured information” (sometimes referred to as “content”) useable. For business information companies, “useable” means that the content can be sold via subscription to a digital product . For enterprises, “useable” means that the content can drive digital process transformation and AI . We work with all classes of data, including sensitive and protected data.

We also develop digital products for business information companies and digital systems which replace legacy systems and processes.

We continued to implement a strategy we initiated in 2017 focused on technology differentiation, increasingly taking an innovation-led approach to create value for clients while driving leaner, more cost-effective operations.

Our Approach: Deep Neural Networks and Experts-in-the-loop

Finding the right approach for mixing machines and experts-in-the-loop for each customer requirement is one of our core competencies. On the human side, we have over three thousand staff with deep domain expertise in legal, tax, regulatory, scientific, technical, and health-related content. Many hold advanced degrees. They work from our global operations centers in North and South America, Europe, Israel, India, Sri Lanka and the Philippines. Our operations centers in Asia are ISO 27001 certified. On the technology side, we use deep learning, a sophisticated machine learning technique which is a branch of artificial intelligence that has enjoyed great success in real-world applications.

In our approach, content is fed first into our deep learning machines that automate much of the work, but distinguish between what they can automatically process and what requires human intervention. Work that requires humans is “pushed” to human experts. Once human experts perform tasks, their output is then fed back to the machines, which, as a result, become “smarter” and achieve over time progressively greater levels of automation. ( See “Our Technology and Infrastructure”, below. )

Our Customers

Our customers include leading digital businesses in banking and financial services, technology, and digital retailing and four of the largest information industry companies in the world, spanning financial, legal, healthcare and scientific vertical markets. Our customers often seek to use digital data in combination with AI to support their businesses or to digitally transform operations. Our information industry customers sell subscriptions to data products that require enhanced digital data.

Our Services

In order to use content in an information product or in AI, there are often significant obstacles that must be overcome: the content may be contained in multiple websites; it may be in disparate formats; and it may lack the necessary semantics and metadata to make it product- or AI-ready. We overcome these obstacles by combining advanced technology and human subject matter experts, enabling us to produce highly refined product- and AI-ready content efficiently and at scale.

Acquiring unstructured data often involves monitoring thousands of sources such as websites in real-time for new or changing content and then automatically extracting the changed content. One of the challenges with web data monitoring and extraction is that the HTML code underlying web pages often changes, causing programs and scripts to break. We overcome this by using enterprise-class tools for content change detection and extraction. These tools use both machine learning-based visual data abstraction combined with traditional programmatic approaches. With visual abstraction, the technology knows what the content “looks like” and recognizes when a data element moves from one spot on a Web page to another.

Digital data transformation can refer to a large number of very specific tasks necessary to create well-structured, valid XML, such as text zoning, editing, format tagging, format conversion, extraction, cross-reference capture, and linking.

Digital data enrichment often includes structural tagging as well as semantic enrichment such as entity tagging and normalization. Data enrichment also includes text categorization - classifying content to systems that organize knowledge such as ontologies and knowledge graphs. This kind of enrichment enables powerful information retrieval and discovery and provides hooks for integration and interoperability.

We provide digital operations management and analytics for an increasing diversity of complex functions. At present, these include IP rights management, contract management, customer relations, data processing, regulatory reporting, publishing workflow management, publisher relationship management, and transaction management. Our services draw on a combination of industry and functional expertise from our thousands of global delivery resources, many with advanced degrees, augmented by sophisticated machine learning and robotics process automation (RPA) technology.

Our digital operations management solutions often leverage our core competency in augmenting human expert effort with advanced technology and configuring the technology to continually improve as a result of expert feedback. In this way, we provide high-quality, cost-effective solutions for our customers that become increasingly automated, resulting in further cost and improvements in both productivity and response time.

Many of the operations we manage are mission-critical for our customers (as opposed to back-office), are unique, and have multiple contingencies and potential points of failure. To manage this complexity, we have developed methodologies and best practices around process management and risk mitigation, and we work with clients at a strategic business and technology level.

We develop and maintain applications used in conjunction with product- or AI-ready content we create. Our content applications address knowledge representation, search/discovery, and digital transformation of customer workflows. Our scope of services spans full-stack development, knowledge engineering, and integration support.

Our Technology and Infrastructure

Our deep learning technologies are built on frameworks such as TensorFlow and Torch that layer algorithms to create artificial neural networks that continuously learn on the job, constantly improving the quality and accuracy of results.

Our content processing application infrastructure consists of three integrated tiers. The first tier is an orchestration layer featuring a console we use to configure our workflow engines. Each workflow is customized around the content acquisition, transformation, and enrichment tasks required to transform a specific input into a specific output. This is where we oversee and manage AI, setting our accuracy thresholds and quality assurance parameters to decide when cognitive tasks that have been performed by machines are sent for human review.

The second tier of our data processing application infrastructure is a microservices layer of deep learning networks which perform discrete tasks automatically. Microservices are invoked by workflows via RESTful APIs. Our deep neural networks are trained to understand domain-specific terms and meaning and can be deployed to enforce privacy and maintain customer-proprietary learning. We have built domain-specific and task-specific microservices that perform deep sequence labelling, text categorization, and computer vision. The technology works by observing text patterns using algorithms and assigning labels based on these observed values.

For each cognitive decision, the machines provide a result together with a confidence score. A high confidence score means the machine is confident that it has performed accurately and the content is ready to be deployed in an information product or for AI. A low confidence score means expert review is required.

Our third layer is human experts-in-the-loop. When an expert review is required, the human experts use tools we refer to as “workbenches” to apply their expertise to a task. The workbenches are integrated with the microservice endpoints in one of two ways. In one deployment, only low-confidence data is sent to workbenches for human review. In an alternative deployment, we send both high-confidence and low-confidence data back to the workbench, displaying the confidence level for each decision taken by the machine. In both of these deployments, human-reviewed work is retroactively fed back into the deep neural network, enabling it to get smarter. The result is continuous, predicable improvement and progressively greater levels of automation.

Our data storage and application hosting platform has been built utilizing an innovative enterprise infrastructure platform enabling robust performance scaling, strong security, high availability, and advanced business continuity. We support a range of strategies to suit our customers’ needs for data security, compliance, scalability and reliability. We can host data and applications in our own data centers at our production facilities or we can utilize third-party cloud services which provide the benefit of “infinite scalability” of hardware resources. When we are processing sensitive information for banks and insurance companies, we utilize U.S.-based, co-located data centers in combination with advanced data encryption (both at rest and in motion to the Advanced Encryption Standard 256 or similar standard) and desktop virtualization technologies, ensuring that data never leaves secured environments in the U.S. We employ a range of security features including monitored firewalls and intrusion detection devices. We can deploy on-premises, as well, and our machine learning microservices can be consumed via API.

We comply with the requirements of the United States Health Insurance Portability and Accountability Act of 1996 as amended (HIPAA) (including by the Health Information Technology for Economic and Clinical Health Data (HITECH)) and the United Kingdom’s Data Protection Act 2018, as applicable. Innodata Inc. is certified to the EU-U.S. Privacy Shield framework, which certification includes Synodex and Agility as covered entities.

Our DDS segment includes our Innodata docGenix, LLC subsidiary (docGenix). As of March 31, 2019, Innodata Inc. owned 94% of docGenix.

Our Venture Businesses

Our venture businesses provide specific industry solutions. We have chosen to invest in these businesses because they leverage one or more of our core capabilities (typically data acquisition/transformation/enrichment at scale, content applications, and global workforce deployment), have attractive business models such as SaaS (software-as-a-service), and expand our addressable market. Each venture business is reported as its own business segment.

Synodex Segment

Synodex enables clients in the insurance and healthcare sectors to transform medical records into useable digital data and to apply technologies to the digital data to augment decision support.

The main focus of the Synodex business is the extraction and classification of data from unstructured medical records in an innovative way to provide improved data service capabilities for insurance underwriting, insurance claims, medical records management, life settlement claims, and clinical trial support services. Synodex has developed and deployed its APS.Extract ^® product for specific use with life insurance underwriting and claims.

Our Synodex segment operates through our Innodata Synodex, LLC subsidiary. As of March 31, 2019, Innodata Inc. owned 92.5% of Innodata Synodex, LLC. More information about our Synodex segment can be found at www.synodex.com .

Agility Segment

Agility provides tools and related professional services that enable public relations (PR) and communications professionals to discover influencers, amplify messages, monitor coverage, and measure the impact of campaigns. Agility has been ranked as the #1 easiest to use media database and as a leader in media and influencer targeting and media monitoring, out-performing significantly larger, more established competitors. ¹ The media intelligence solutions market is estimated to be $3.2 billion and growing 7% annually. ² The market is highly fragmented, with only a limited number of providers of integrated solutions such as Agility.

Agility’s software-as-a-service (SaaS) tools include:

¹ G2 Crowd

² Burton-Taylor

Agility’s professional services include:

To provide our services, we maintain a big data engine that stores and indexes media content. We index approximately two billion media items each year.

Bulldog Reporter, a publisher of PR-related news and a popular e-newsletter, and the Bulldog Awards, a PR awards program that recognizes outstanding performance among PR and communications professionals and agencies, are properties of Agility. More information about our Agility segment can be found at www.agilitypr.com

Revenues

Revenue Recognition – commencing January 1, 2018 we recognize revenue in accordance with Accounting Standards Update (“ASU”) 2014-09, Revenue from Contracts with Customers . Revenue is recognized when control of the promised services is transferred to a customer, in an amount that reflects the consideration that we expect to receive in exchange for those services as per the agreement with the customer. In case there are agreements with multiple performance obligations, we identify each performance obligation and evaluate whether the performance obligations are distinct within the context of the agreement at the agreement’s inception. Performance obligations that are not distinct at agreement inception are combined. We allocate the transaction price to each distinct performance obligation proportionately based on the estimated standalone selling price for each performance obligation, if any, and then evaluate how the services are transferred to the customer to determine the timing of revenue recognition.

For the DDS segment, revenue is recognized primarily based on the quantity delivered or resources utilized in the period in which services are performed and performance conditions are satisfied as per the agreement. Revenues for agreements billed on a time-and-materials basis are recognized as services are performed. Revenues under fixed-fee agreements, which are not significant to the overall revenues, are recognized based on the proportional performance method of accounting, as services are performed, or milestones are achieved.

For the Synodex segment, revenue is recognized primarily based on the quantity delivered in the period in which services are performed and performance conditions are satisfied as per the agreement. A portion of our Synodex segment revenue is derived from licensing our functional software and providing access to our hosted software platform. Revenue from such services is recognized monthly when access to the service is provided to the end user; all parties to the agreement have agreed to the agreement; each party’s rights are identifiable; the payment terms are identifiable; the agreement has commercial substance; and collection is probable.

The Agility segment derives its revenue primarily from subscription arrangements and provision of enriched media analysis services. It also derives revenue as a reseller of corporate communication solutions. Revenue from subscriptions is recognized monthly when access to the service is provided to the end user; all parties to the agreement have agreed to the agreement; each party’s rights are identifiable; the payment terms are identifiable; the agreement has commercial substance; and collection is probable. Revenue from enriched media analysis services is recognized when the services are performed, and performance conditions are satisfied. Revenues from the reseller agreements, which are not significant to the overall revenues, are recognized at gross with our functioning as a principal due to our meeting the following criteria. We act as the primary obligor in the sales transaction; assume the credit risk; set the price; can select suppliers; and are involved in the execution of the services, including after sales service.

Revenues include reimbursement of out-of-pocket expenses, with the corresponding out-of-pocket expenses included in direct operating costs.

We consider U.S. GAAP criteria for determining whether to report revenue gross as a principal versus net as an agent.  Factors considered include whether we are the primary obligor, have risks and rewards of ownership, and bear the risk that a customer may not pay for the services performed.  If there are circumstances where the above criteria are not met and therefore, we are not the principal in providing services, amounts received from customers are presented net of payments in the condensed consolidated statements of operations and comprehensive loss.

Contract acquisition cost for our Agility segment is amortized over the term of the subscription agreement which normally has a duration of 12 months or less. We review these costs on a periodic basis to determine the need to adjust the carrying values for pre-terminated contracts.

Direct Operating Costs

Direct operating costs consist of direct payroll, occupancy costs, data center hosting fees, content acquisition costs, depreciation and amortization, travel, telecommunications, computer services and supplies, realized gain (loss) on forward contracts, foreign currency revaluation gain (loss), and other direct expenses that are incurred in providing services to our clients.

Selling and Administrative Expenses

Selling and administrative expenses consist of management and administrative salaries, sales and marketing costs including commissions, new services research and related software development, third-party software, advertising and trade conferences, professional fees and consultant costs, and other administrative overhead costs.

Valuation of Goodwill and Intangible Assets

The Company performs valuations of assets acquired and liabilities assumed on each acquisition accounted for as a business combination and allocates the purchase price of each acquired business to its respective net tangible and intangible assets and liabilities. Acquired intangible assets principally consist of technology, customer relationships, backlog and trademarks. Liabilities related to intangibles principally consist of unfavorable vendor contracts. The Company determines the appropriate useful life by performing an analysis of expected cash flows based on projected financial information of the acquired businesses. Intangible assets are amortized over their estimated useful lives using the straight-line method, which approximates the pattern in which the majority of the economic benefits are expected to be consumed. Intangible liabilities are amortized into direct operating costs ratably over their expected related revenue streams over their useful lives.

Goodwill represents the excess of the cost of an acquired entity over the fair value of the acquired net assets. The Company does not amortize goodwill but evaluates it for impairment at the reporting unit level annually during the third quarter of each fiscal year (as of September 30 of that quarter) or when an event occurs, or circumstances change, that indicates the carrying value may not be recoverable. In 2018 the Company adopted ASU 2017-04, Intangibles - Goodwill and Other (Topic 350): Simplifying the Accounting for Goodwill Impairment. Under the newly adopted guidance, the optional qualitative assessment, referred to as “Step 0”, and the first step of the quantitative assessment (“Step 1”) remained unchanged versus the prior guidance. However, the requirement to complete the second step (“Step 2”), which involved determining the implied fair value of goodwill and comparing it to the carrying value of that goodwill to measure the impairment loss, was eliminated. As a result, Step 1 will be used to determine both the existence and amount of goodwill impairment. An impairment loss will be recognized for the amount by which the reporting unit’s carrying amount exceeds its fair value, not to exceed the carrying value of goodwill in that reporting unit.

Adjusted EBITDA

In addition to measures of financial performance presented in our consolidated financial statements, we monitor “Adjusted EBITDA” to help us evaluate our ongoing operating performance including our ability to operate the business effectively.

We define Adjusted EBITDA as net income (loss) attributable to Innodata Inc. and its subsidiaries in accordance with U.S. GAAP before income taxes, depreciation, amortization of intangible assets, impairment charges, changes in fair value of contingent consideration, stock-based compensation, loss attributable to non-controlling interests and interest income (expense).

We believe Adjusted EBITDA is useful to our management and investors in evaluating our operating performance and for financial and operational decision-making purposes. In particular, Adjusted EBITDA facilitates comparisons of the core operating performance of our Company from period to period on a consistent basis and helps us identify underlying trends in our business. We believe it provides useful information about our operating results, enhances the overall understanding of our past performance and future prospects and allows for greater transparency with respect to key metrics used by the management in our financial and operational decision-making. We use this measure to establish operational goals for managing our business and evaluating our performance. 

Adjusted EBITDA has limitations as an analytical tool and should not be considered in isolation or as a substitute for results reported under U.S. GAAP. Some of these limitations are:

Adjusted EBITDA should be considered as a supplement to, and not as a substitute for or superior to, U.S. GAAP net income.

The following tables reconcile net income (loss) to Adjusted EBITDA for the periods presented (in thousands):

Results of Operations

Three Months Ended March 31, 2019 and 2018

Revenues

Total revenues were $13.7 million for the three months ended March 31, 2019 compared to $14.1 million for the three months ended March 31, 2018, a decrease of approximately $0.4 million or 3%.

Revenues from the DDS segment were $10.2 million and $10.5 million for the three months ended March 31, 2019 and 2018, respectively, a decrease of approximately $0.3 million or 3%. The decrease is attributable to a reduction in volume from one client offset in part by volume increases from other clients.

Revenues from the Synodex segment were $1.0 million for each of the three months ended March 31, 2019 and 2018.

Revenues from the Agility segment were $2.5 million and $2.6 million for the three months ended March 31, 2019 and 2018, respectively, a decrease of approximately $0.1 million or 4%.

Two clients in the DDS segment generated an aggregate of approximately 26% of the Company’s total revenues for the three months ended March 31, 2019 and 32% of the Company’s total revenues for the three months ended March 31, 2018. No other client accounted for 10% or more of total revenues during these periods. Further, revenues from non-U.S. clients accounted for 52% and 58% of the Company’s total revenues for the three months ended March 31, 2019 and 2018, respectively.

Direct Operating Costs

Direct operating costs were $9.6 million and $9.9 million for the three months ended March 31, 2019 and 2018, respectively, a decrease of $0.3 million or 3%. Direct operating costs as a percentage of total revenues were 70% for each of the three months ended March 31, 2019 and March 31, 2018.

Direct operating costs for the DDS segment were approximately $7.3 million and $7.5 million for the three months ended March 31, 2019 and 2018, respectively, a net decrease of approximately $0.2 million or 3%.  The decrease in direct operating costs is primarily attributable to lower labor costs of $0.3 million and reduced delivery center costs of $0.4 million, offset in part by $0.5 million of foreign exchange losses. Direct operating costs for the DDS segment as a percentage of DDS segment revenues were 71% for each of the three months ended March 31, 2019 and 2018, respectively.

Direct operating costs for the Synodex segment were $0.7 million for the three months ended March 31, 2019 and $0.8 million for the three months ended March 31, 2018, net of intersegment profits, a decrease of $0.1 million or 12%. The decline in direct operating costs reflects production efficiencies realized. Direct operating costs for the Synodex segment as a percentage of Synodex segment revenues were 71% and 80% for the three months ended March 31, 2019 and 2018, respectively.

Direct operating costs for the Agility segment were $1.6 million for the three months ended March 31, 2019 and 2018, respectively. Direct operating costs for the Agility segment as a percentage of Agility segment revenues were 64% and 62% for the three months ended March 31, 2019 and 2018, respectively.

Selling and Administrative Expenses

Selling and administrative expenses were $4.6 million for the three months ended March 31, 2019 compared to $3.9 million for the three months ended March 31, 2018, an increase of $0.7 million or approximately 18%. Selling and administrative expenses as a percentage of total revenues were 34% and 28% for the three months ended March 31, 2019 and 2018, respectively.

Selling and administrative expenses for the DDS segment were $3.0 million and $2.3 million for the three months ended March 31, 2019 and 2018, respectively, an increase of $0.6 million or 26%. The increase in selling and administrative expenses are primarily attributable to higher professional fees, incentives and new hires. As a percentage of DDS revenues, DDS selling and administrative expenses were 29% and 22% for the three months ended March 31, 2019 and March 31, 2018, respectively.

Selling and administrative expenses for the Synodex segment were $0.1 million for the three months ended March 31, 2019 and $0.2 million for the three months ended March 31, 2018, respectively, a decrease of $0.1 million or 50%. Selling and administrative expenses for the Synodex segment as a percentage of Synodex segment revenues were 13% for the three months ended March 31, 2019 and 20% for the three months ended March 31, 2018.

Selling and administrative expenses for the Agility segment were $1.5 million and $1.4 million for the three months ended March 31, 2019 and 2018, respectively, a decrease of $0.1 million or 7%. Selling and administrative expenses for the Agility segment as a percentage of Agility segment revenues were 61% for the three months ended March 31, 2019 and 54% for the three months ended March 31, 2018.

Income Taxes

We recorded a provision for income taxes of ($28,000) and $0.6 million for the three months ended March 31, 2019 and 2018, respectively. Taxes primarily consist of a provision for foreign taxes recorded in accordance with the local tax regulations by our foreign subsidiaries. Effective income tax rates are disproportionate due to the losses incurred by our U.S. entity and our Canadian subsidiaries and a valuation allowance recorded on deferred taxes on these entities.

The reconciliation of the U.S. statutory rate with the Company’s effective tax rate for the three-month period ended March 31, 2019 is summarized as below:

Despite access to overseas earnings and the resulting toll charge, we intend to indefinitely reinvest foreign earnings in our foreign subsidiaries on account of the foreign jurisdiction withholding taxes that we would have to incur on the actual remittances. Unremitted earnings of foreign subsidiaries amounted to approximately $20.8 million at March 31, 2019. If such earnings are repatriated in the future, or are no longer deemed to be indefinitely reinvested, we would have to accrue the applicable amount of foreign jurisdiction withholding taxes associated with such remittances.

We have a valuation allowance on all of our U.S. deferred tax assets on account of continuing losses incurred by our U.S. entity. In addition, we also have a valuation allowance on the deferred tax assets of our Canadian subsidiaries. Our Canadian subsidiaries also have research and development expenditures available to reduce taxable income in future years, which may be carried forward indefinitely. The potential benefits from these balances have not been recognized for financial statement purposes.

The Company is subject to Federal income tax, as well as income tax in various states and foreign jurisdictions. The Company has open periods for US Federal and state taxes from 2015 through 2018. Various foreign subsidiaries currently have open tax years from 2003 through 2018. Refer to Item 1, Note 3 for more information about Tax Assessments.

Net Loss

We incurred a net loss of $0.5 million during the three months ended March 31, 2019 compared to a net loss of $0.3 million during the three months ended March 31, 2018.

Net income for the DDS segment was break-even for the three months ended March 31, 2019 and March 31, 2018, net of intersegment profits. Net income for the Synodex segment was $150,000 for the three months ended March 31, 2019 compared to break-even for the three months ended March 31, 2018, net of intersegment profits. Net loss for the Agility segment was $650,000 and $300,000 for the three months ended March 31, 2019 and 2018, respectively.

Adjusted EBITDA

Adjusted EBITDA for the three months ended March 31, 2019 was $0.5 million compared to $1.3 million for the three months ended March 31, 2018. Adjusted EBITDA for the DDS segment was $0.6 million and $1.2 million for the three months ended March 31, 2019 and 2018, respectively. Adjusted EBITDA for the Synodex segment was $0.2 million and break-even for the three months ended March 31, 2019 and 2018, respectively. Adjusted EBITDA for the Agility segment was a net loss of $0.3 million and income of $0.1 million for the three months ended March 31, 2019 and 2018, respectively.

At March 31, 2019, we had cash and cash equivalents of $12.2 million, of which $5.7 million was held by our foreign subsidiaries, and the $6.5 million balance was held in the United States. Based on our calculations, as a result of the one-time transition tax on the deemed repatriation of post-1986 undistributed foreign subsidiary earnings and profits, we will not have any liability for taxes with respect to repatriated foreign earnings under Internal Revenue Service Regulation 956 until our cumulative exposure for unrepatriated foreign earnings reaches a threshold of $25.8 million. Although we do not currently intend to repatriate amounts held by foreign subsidiaries, we could if needed repatriate these amounts less up to 21% in withholding taxes of foreign jurisdictions.

We have used, and plan to use, our cash and cash equivalents for (i) investments in the Agility segment; (ii) the expansion of our other operations; (iii) technology innovations; (iv) product management and strategic marketing; (v) general corporate purposes, including working capital; and (vi) possible business acquisitions. As of March 31, 2019, we had working capital of approximately $11.1 million, as compared to working capital of approximately $13.0 million as of December 31, 2018.

We did not have any material commitments for capital expenditure as of March 31, 2019.

We believe that our existing cash and cash equivalents and internally generated funds will provide sufficient sources of liquidity to satisfy our financial needs for the next 12 months. However, we have no debt facilities or lines of credit, and reductions in our cash and cash equivalents from operating losses, capital expenditures, adverse legal decisions, acquisitions or otherwise would materially and adversely affect the Company.

Cash Flows

Net Cash Provided by Operating Activities

Cash provided by our operating activities for the three months ended March 31, 2019 was $2.1 million primarily on account of the $2.5 million decrease in our accounts receivable and $0.1 million net decrease in other working capital accounts. This favorable result was reduced by our net loss of $0.5 million in the quarter.  The reduction in accounts receivable  is a result of improvements in our collection effort during the period.  Refer to the condensed consolidated statement of cash flows for further details.

Our days’ sales outstanding (DSO) were 62 days for the three months ended March 31, 2019 and 66 days for the year ended December 31, 2018. We calculate DSO for a reported period by first dividing the total revenues for the period by the average net accounts receivable for the period (which is the sum of the net accounts receivable at the beginning of the period and the net accounts receivable at the end of the period, divided by two), to yield an amount we refer to as the “accounts receivable turnover.” Then we divide the total number of days within the reported period by the accounts receivable turnover to yield DSO expressed in number of days.

Net Cash Used in Investing Activities

For the three months ended March 31, 2019, cash used in our investing activities was $0.5 million. These expenditures principally consisted of purchases of technology equipment including servers, network infrastructure and workstations.

During the next 12 months, we anticipate that capital expenditures for ongoing technology, equipment and infrastructure upgrades will approximate $1.0 million to $2.0 million, a portion of which we may finance.

Net Cash Used in Financing Activities

For the three months ended March 31, 2019, cash used in financing activities was for payments of long-term obligations of $0.4 million.

Inflation, Seasonality and Prevailing Economic Conditions

Our most significant costs are the salaries and related benefits of our employees in Asia. We are exposed to higher inflation in wage rates in the countries in which we operate. We generally perform work for our clients under project-specific contracts, requirements-based contracts or long-term contracts. We must adequately anticipate wage increases, particularly on our fixed-price contracts. There can be no assurance that we will be able to recover cost increases through increases in the prices that we charge for our services to our clients.

Our quarterly operating results are subject to certain fluctuations. We experience fluctuations in our revenue and earnings as we replace and begin new projects, which may have some normal start-up delays, or we may be unable to replace a project entirely. These and other factors may contribute to fluctuations in our operating results from quarter to quarter. In addition, as some of our Asian facilities are closed during holidays in the fourth quarter, we typically incur higher wages, due to overtime, that reduce our margins.

Our Synodex subsidiary experiences seasonal fluctuations in revenues. Typically, revenue is lowest in the third quarter of the calendar year and highest in the fourth quarter of the calendar year. The seasonality is directly linked to the number of life insurance applications received by the insurance companies.

Critical Accounting Policies and Estimates

Our discussion and analysis of our results of operations, liquidity and capital resources is based on our consolidated financial statements which have been prepared in conformity with accounting principles generally accepted in the United States of America. The preparation of these consolidated financial statements requires us to make estimates and judgments that affect the reported amounts of assets, liabilities, revenues and expenses, and disclosure of contingent assets and liabilities. On an ongoing basis, we evaluate our estimates and judgments, including those related to revenue recognition, allowance for doubtful accounts and billing adjustments, long-lived assets, intangible assets, goodwill, valuation of deferred tax assets, value of securities underlying stock-based compensation, litigation accruals, pension benefits, purchase price allocation of Agility, valuation of derivative instruments and estimated accruals for various tax exposures. We base our estimates on historical and anticipated results and trends and on various other assumptions that we believe are reasonable under the circumstances, including assumptions as to future events. These estimates form the basis for making judgments about the carrying values of assets and liabilities that are not readily apparent from other sources. By their nature, estimates are subject to an inherent degree of uncertainty. Actual results may differ from our estimates and could have a significant adverse effect on our consolidated results of operations and financial position. We believe the following critical accounting policies affect our more significant estimates and judgments in the preparation of our consolidated financial statements.

The significant accounting policies used in preparing these condensed consolidated financial statements are the same as those described in the Company’s Annual Report on Form 10-K, unless otherwise noted.

Recent Accounting Pronouncements

In January 2016, the Financial Accounting Standards Board (FASB) issued ASU No. 2016-01, Financial Instruments – Overall (“Subtopic 825-10”): Recognition and Measurement of Financial Assets and Financial Liabilities (“ASU 2016-01”), which updates certain aspects of recognition, measurement, presentation and disclosure of financial instruments. The Company adopted this standard in the first quarter of 2019, and it did not have a material impact on the Company’s consolidated financial statements.

In August 2018, the FASB issued ASU 2018-14, “Compensation-Retirement Benefits-Defined Benefits Plans-General: Disclosure Framework-Changes to the Disclosure Requirements for Defined Benefit Plans” (“ASU 2018-14”) that makes minor changes to the disclosure requirements for employers that sponsor defined benefit pension and/or other postretirement benefit plans. The guidance eliminates requirements for certain disclosures that are no longer considered cost beneficial and adds new disclosure requirements that the FASB considers pertinent. ASU 2018-14 is effective for fiscal years ending after December 15, 2020 for public entities; early adoption is permitted. We are currently evaluating the early adoption of ASU-2018-14 but do not expect it to have a material impact to the Company’s consolidated financial statements.

Off-Balance Sheet Arrangements

None.