APPEN LIMITED — AGM Information 2021

May 27, 2021

==> picture [114 x 27] intentionally omitted <==

Appen Limited Level 6, 9 Help Street Chatswood NSW 2067

Tel : 02 9468 6300 www.appen.com ACN: 138 878 298

ASX ANNOUNCEMENT

28 May 2021

2021 ANNUAL GENERAL MEETING – CEO’S PRESENTATION

Appen Limited ( Appen ) (ASX: APX) provides the attached CEO’s presentation to be delivered at today’s Annual General Meeting commencing at 10.00am AEST.

The webcast of the AGM can be viewed at: www.agmlive.link/APX21.

Authorised for release by the Board.

Please contact for more information:

Linda Carroll Investor Relations +61 2 9468 6300 investorrelations@appen.com www.appen.com/investors

About Appen

Appen collects and labels images, text, speech, audio, video, and other data used to build and continuously improve the world’s most innovative artificial intelligence systems. Our expertise includes having a global crowd of over 1 million skilled contractors who speak over 235 languages, in over 70,000 locations and 170 countries, and the industry’s most advanced AI-assisted data annotation platform. Our reliable training data gives leaders in technology, automotive, financial services, retail, healthcare, and governments the confidence to deploy world-class AI products. Founded in 1996, Appen has customers and offices globally.

==> picture [888 x 154] intentionally omitted <==

==> picture [115 x 28] intentionally omitted <==

Appen Limited Annual General Meeting CEO Presentation

==> picture [88 x 22] intentionally omitted <==

Disclaimer

==> picture [103 x 70] intentionally omitted <==

The forward-looking statements included in these materials involve subjective judgement and analysis and are subject to significant uncertainties, risks, contingencies, many of which are outside the control of, and are unknown to Appen Limited. In particular, they speak only as of the date of these materials, they are based on particular events, conditions or circumstances stated in the materials, they assume the success of Appen Limited’s business strategies, and they are subject to significant regulatory, business, competitive, currency and economic uncertainties and risks.

Appen Limited disclaims any obligation or undertaking to disseminate any updates or revisions to any forward-looking statements in these materials to reflect any change in expectations in relation to any forward-looking statements or any change in events, conditions or circumstances on which any such statement is based. You should monitor any announcements by the company lodged with the ASX. Nothing in these materials shall under any circumstances create an implication that there has been no change in the affairs of Appen Limited since the date of these materials. Organisation structure is subject to change.

No representation, warranty or assurance (express or implied) is given or made in relation to any forward-looking statement by any person (including Appen Limited). In particular, no representation, warranty or assurance (express or implied) is given in relation to any underlying assumption or that any forward-looking statement will be achieved. Actual future events and conditions may vary materially from the forward-looking statements and the assumptions on which the forward-looking statements are based. Given these uncertainties, readers are cautioned to not place undue reliance on such forward-looking statements. To the maximum extent permitted by law, Appen disclaims all liability and responsibility (including without limitation, any liability arising from fault or negligence) for any direct or indirect loss or damage which may arise or be suffered through use or reliance on anything contained in, or omitted from, this presentation.

Information in this presentation should be read in conjunction with Appen’s latest and prior interim and annual reports, and Appen’s announcements via the ASX.

==> picture [88 x 22] intentionally omitted <==

2

==> picture [960 x 82] intentionally omitted <==

Artificial intelligence...

Appen makes AI work in the real world by delivering high-quality training data at scale.

Our clients include the world’s largest technology companies, global leaders in automotive, financial services, retail, healthcare, and government agencies.

...informed by Appen.

==> picture [89 x 21] intentionally omitted <==

3

Solid growth maintained in FY20

A$	FY2020	vs FY2019
Group revenue	$599.9M	+12%
Relevance	$538.2M	+15%
Speech & Image	$61.2M	-10%
Underlying EBITDA1	$108.6M	+8%
Underlying EBITDA margin	18.1%	vs 18.8%
Dividend per share	10.0c	+11%

==> picture [103 x 70] intentionally omitted <==

High growth and performance in 1H20

Moderated in 2H20 due to strong Australian dollar and COVID-19 impacts including:

• impact of lower digital ad revenue and uncertain outlook on customer spend

• deferral and reprioritisation of projects

• restricted face-to-face sales and customer engagement

Strong balance sheet - A$78M in cash and no debt as at 31 December 2020

Final dividend of 5.5cps and interim dividend of 4.5cps, both 50% franked

1. Underlying EBITDA excludes transaction costs, acquisition related share based payment expenses and fair value adjustments (consideration adjustments) for the Figure Eight acquisition.

==> picture [88 x 22] intentionally omitted <==

4

New customer and project wins strengthen our position

==> picture [103 x 70] intentionally omitted <==

==> picture [112 x 13] intentionally omitted <==

----- Start of picture text -----

Customer wins
----- End of picture text -----

==> picture [254 x 164] intentionally omitted <==

----- Start of picture text -----

46
36
30
24
Q1 Q2 Q3 Q4
----- End of picture text -----

==> picture [463 x 200] intentionally omitted <==

----- Start of picture text -----

Project wins China revenue
A$M
2.8
+60%
QoQ
1.9
+34 [%]
1.3
in 2020
0.7
Q1 Q2 Q3 Q4
----- End of picture text -----

==> picture [93 x 15] intentionally omitted <==

----- Start of picture text -----

Project wins
----- End of picture text -----

• 136 new customers in 2020

• Multiple industries, geographies, data modalities and use cases

• Growth enabled by sales and marketing investment and annotation platform capabilities

• 34% increase in projects with existing major customers

• Customer use of our annotation platform is growing, enabling more data types and use cases, tighter integration and greater retention

 China revenue growth 60% quarter on quarter

• Customers include major Chinese technology players plus autonomous vehicle, health and education companies

• Capabilities aligned with customers’ new product development priorities

==> picture [88 x 22] intentionally omitted <==

5

Ongoing growth in committed revenue

==> picture [103 x 70] intentionally omitted <==

Annual contract value (ACV)

==> picture [290 x 194] intentionally omitted <==

----- Start of picture text -----

At end of period
US$M
124.4
103.0
98.7
25.0
2H 2019 1H 2020 2H 2020 1 Feb 2021
----- End of picture text -----

==> picture [364 x 212] intentionally omitted <==

----- Start of picture text -----

Revenue by type
A$M
100% 97% 93% 88% 69%
Project
Committed
31%
3%
12%
7%
2H 2018 1H 2019 2H 2019 1H 2020 2H 2020
----- End of picture text -----

• 399% increase in ACV to US$124.4M (1 Feb 21 v 2H19)

• 343% increase in committed revenue (2H20 v 2H19)

• Increase underpinned by expansion of enterprisewide platform agreement with existing major customer

• A$92.0M of committed revenue in 2H20, 31% of total – up from A$36.3M in 1H20, 12% of total

• Some smaller customers continued to be impacted by COVID-19 in 2H20

==> picture [88 x 22] intentionally omitted <==

6

Data remains a major obstacle for AI

==> picture [103 x 70] intentionally omitted <==

In a recent survey, 18% of respondents see lack of data or data quality as the major bottleneck to AI

Many AI practitioners spend most of their time collecting and preparing data

==> picture [406 x 297] intentionally omitted <==

“… Airbnb have discovered that nearly 70% of the time a data scientist spends developing machine learning models – is not doing the actual modelling, but collecting data and feature engineering .”

https://read.hyperight.com/ml-powered-accommodation-hunt-airbnb-approach/

==> picture [88 x 22] intentionally omitted <==

O’Reilly: AI Adoption in the Enterprise 2021. Survey n = 3,574

7

Shift from “model-centric” to “data-centric” AI

==> picture [313 x 318] intentionally omitted <==

==> picture [6 x 8] intentionally omitted <==

----- Start of picture text -----

8
----- End of picture text -----

==> picture [88 x 22] intentionally omitted <==

May 27, 2021

Source: Landing.ai

Automation critical to the labelling process

==> picture [103 x 70] intentionally omitted <==

==> picture [67 x 60] intentionally omitted <==

Pre-labelling

• AI performs an initial ‘best guess’ of the annotation

• Crowd workers check and correct the pre-label (if required)

• Significantly reduces annotation time

==> picture [67 x 58] intentionally omitted <==

Speed labelling

• AI models that assist crowd workers by automating slow and/or manual tasks

• Works similar to an autocomplete function

• Significantly reduces annotation time

==> picture [67 x 61] intentionally omitted <==

Smart Validators

• AI models that verify crowd output before they are submitted

• Crowd get notified if input is not within expected thresholds

• Improves data quality and overall worker performance

Appen’s product suite includes many automation features

==> picture [88 x 22] intentionally omitted <==

9

==> picture [103 x 70] intentionally omitted <==

Real-world models typically use a combination of data-sources and techniques

Contribution from different model/data approaches (Illustrative)

==> picture [791 x 303] intentionally omitted <==

----- Start of picture text -----

Supervised
Supervised
US English chatbot Self-supervised Supervised (Human
Transfer learning (AI-assisted human
for a bank learning (Off the shelf data) annotated
annotated data)
data)
Self- Supervised
French chatbot for Transfer Supervised Supervised
supervised (AI-assisted human
a bank learning (Off the shelf data) (Human annotated data)
learning annotated data)
Qatari Arabic Self- Supervised
Transfer Supervised Supervised
chatbot for a marine supervised (AI-assisted human annotated
learning (Off the shelf data) (Human annotated data)
insurance company learning data)
----- End of picture text -----

Appen AI training data solutions

==> picture [88 x 22] intentionally omitted <==

10

Popular AI techniques rely on human involvement

==> picture [103 x 70] intentionally omitted <==

AI technologies used in mature practices

Techniques that typically require some level of human annotation and/or preparation

Source: O’Reilly: AI Adoption in the Enterprise 2021. Survey n = 3,574

==> picture [88 x 22] intentionally omitted <==

11

Appen continues to evolve

==> picture [103 x 70] intentionally omitted <==

	From	To
Data type	Language data	AI data
Delivery model	Service led	Product led
Revenue	Project based	Committed
Customer	Concentration	Diversification
Org structure	Functional alignment	Customer alignment
Reporting	Data modality, AUD	Strategy led, USD

==> picture [88 x 22] intentionally omitted <==

12

==> picture [240 x 305] intentionally omitted <==

Our areas of focus

Our future is product led, enabling us to deliver high-quality Product training data, faster, at larger scale, with improved unit Led economics, and is a foundation for future capabilities

Customer We are aligning our operations to better support the needs Centric of our target customer cohorts

==> picture [88 x 22] intentionally omitted <==

13

Appen’s high value product suite

==> picture [103 x 70] intentionally omitted <==

==> picture [861 x 397] intentionally omitted <==

----- Start of picture text -----

New products
Appen Appen Data Appen Appen Appen
Connect Annotation Platform Intelligence In-Platform Audit Mobile
Match our global Collect and annotate Powers Appen products Organise and analyse Engages, enables and
crowd workforce to training data with proprietary ML models training data to identify expands crowd
annotation tasks quality, distribution & bias
AI-augmented data In-built crowd
AI-enabled crowd Native integration
collection and Embedded expertise management
management with our crowd
labelling features
Improves speed, Increases internal Creates
Delivers high Reduces risk for
quality, scale and productivity and competitive
quality annotation customers
unit economics crowd experience differentiation
----- End of picture text -----

==> picture [88 x 22] intentionally omitted <==

14

Customer-centric org structure

==> picture [103 x 70] intentionally omitted <==

• Four customer focused business units with end-to-end operational and P&L accountability

• Respond quickly to changing customer needs and market trends

• We are optimising resources to align with the new operating model and technologydriven productivity benefits

• Reduced delivery resource requirements will deliver benefits from 2H21

==> picture [526 x 355] intentionally omitted <==

----- Start of picture text -----

Enterprise
Govt
Global
China
Crowd and
Product Engineering Corporate
HR
----- End of picture text -----

==> picture [88 x 22] intentionally omitted <==

15

Revenue by customer-focused business units

==> picture [103 x 70] intentionally omitted <==

==> picture [288 x 224] intentionally omitted <==

==> picture [277 x 168] intentionally omitted <==

Global customers deliver solid growth underpinned by large programs

Enterprise, Government and China provide high growth and will improve customer distribution over time

==> picture [88 x 22] intentionally omitted <==

Note: All figures in US$M. 1. Annualised CAGR from H1 2019 to H2 2020.

16

Segmenting by offering shows very high growth in New Markets

==> picture [103 x 70] intentionally omitted <==

==> picture [249 x 184] intentionally omitted <==

==> picture [258 x 189] intentionally omitted <==

Services provided to leading US tech companies utilising their data annotation tools

Revenue from leading US tech companies through Appen products and Enterprise, Government and China

==> picture [88 x 22] intentionally omitted <==

Note: All figures in US$M. 1. Annualised CAGR from H1 2019 to H2 2020.

17

Solid earnings from Global Services supports New Markets growth

==> picture [247 x 223] intentionally omitted <==

==> picture [251 x 190] intentionally omitted <==

Global Services provides solid earnings with margin expansion through scale and technology

Ongoing investments into New Markets to drive growth

==> picture [88 x 22] intentionally omitted <==

Note: All figures in US$M. EBITDA includes all corporate costs excluding share-based payment and FX gains and losses.

18

2021 outlook

2021 revenue growth outlook

• Global Services skewed to H2; full-year mid to high single-digit growth

• New Markets expected growth circa 25%, in line with broader AI market

==> picture [103 x 70] intentionally omitted <==

Resource optimisation

• Restructure and tech-enabled productivity allow resource optimisation, mainly in delivery resources

• Restructuring costs in H1 2021, partial benefits to be realised in H2 2021

• Full year gross cost savings (before reinvestment) of US$15M expected in 2022

==> picture [88 x 22] intentionally omitted <==

19

Trading update

==> picture [103 x 70] intentionally omitted <==

• Year-to-date revenue plus orders in hand of ~US$260M[1] at end of April 2021, consistent with prior methodology[2]

• Heavy weighting to underlying EBITDA in 2H21 due to:

• Key projects that were delayed in late 2020 are returning with a skew to delivery in 2H21

• 1H21 cost base reflects full year cost of 2020 hiring

• Resource optimisation benefits to flow 2H21

• Full year underlying EBITDA guidance of US$83M - US$90M is maintained

1. ~AU$340M at AUD/USD FX rate of 77c. 2. Year-to-date revenue and orders in hand at the same time in 2020 was approximately US$240 million.

==> picture [88 x 22] intentionally omitted <==

20

Appen is accelerating its transformation into an AI powered provider of AI data and solutions

==> picture [240 x 305] intentionally omitted <==

==> picture [88 x 22] intentionally omitted <==

21

==> picture [960 x 82] intentionally omitted <==

==> picture [373 x 373] intentionally omitted <==

Thank you

==> picture [32 x 32] intentionally omitted <==

==> picture [31 x 32] intentionally omitted <==

==> picture [31 x 32] intentionally omitted <==

==> picture [31 x 32] intentionally omitted <==

==> picture [88 x 22] intentionally omitted <==

22