case_study at position 10–15%

61 slides where a page of type case study falls in the 10–15% range of its deck.

McKinsey 2018 p.135/1274 10%
Lebanon Economic Vision
“Industrial construction technologies have been successfully leveraged for housing across the world...”
BondCap p.41/356 11%
2017 Bond Cap Internet Trends 2017
“...Brands + Influencers = Re-Distribution Driving Engagement”
Influencers = Can Impact Followers
BondCap p.44/340 12%
2025 Bond Cap Artificial Intelligence AI 2025
“AI Performance = Increasingly Realistic Conversations Simulating Human Behaviors”
During the test, participants incorrectly identified the left image (Witness A) as human with 87% certainty, saying 'A had human vibes. B had human imitation vibes.' A was actually AI-generated; B was
87% · Turing Test success rate
BondCap p.45/340 13%
2025 Bond Cap Artificial Intelligence AI 2025
“AI Performance = Increasingly Realistic Image Generation...”
BondCap p.46/340 13%
2025 Bond Cap Artificial Intelligence AI 2025
“...AI Performance = Increasingly Realistic Image Generation”
BondCap p.47/340 13%
2025 Bond Cap Artificial Intelligence AI 2025
“AI Performance = Increasingly Realistic Audio Translation / Generation...”
In just two years, ElevenLabs' millions of users have generated 1,000 years of audio content and the company's tools have been adopted by employees at over 60% of Fortune 500 companies.
20M · Monthly Global Site Visits
BondCap p.48/340 13%
2025 Bond Cap Artificial Intelligence AI 2025
“...AI Performance = Evolving to Mainstream Realistic Audio Translation / Generation”
Spotify begins accepting audiobooks AI-translated into 29 languages from ElevenLabs.
678MM · Monthly Active Users
BondCap p.36/334 10%
2019 Bond Cap Internet Trends 2019
“Stitch Fix = Recommendations Drive Growth”
What's really special about Stitch Fix is that 100% of what we sell is based on recommendations.
AirStreetCapital p.38/313 11%
2025 Air Street Capital The State of AI Report 2025
“The Muon Optimizer: expanding the compute-time Pareto frontier beyond AdamW”
Muon demonstrates better data efficiency at large batch sizes, enabling faster training with more devices.
10-15% · Training Time
AirStreetCapital p.39/313 12%
2025 Air Street Capital The State of AI Report 2025
“Cutting your losses: significant memory reduction for LLM training”
CCE achieves a remarkable 24 reduction in memory consumption, taking Gemma 2's loss computation from 24GB down to just 1 MB, while actually running ~5% faster than the best existing methods.
24x · Memory consumption
AirStreetCapital p.41/313 12%
2025 Air Street Capital The State of AI Report 2025
“Learning from superintelligence: AlphaZero teaches chess grandmasters new concepts”
This proof-of-concept suggests a very exciting potential paradigm where superhuman AI systems could become "teachers" rather than just "tools", and help advance human knowledge in domains beyond chess
0.85 · improvement in puzzles solved
AirStreetCapital p.42/313 13%
2025 Air Street Capital The State of AI Report 2025
“Kimi K2: stable trillion-scale MoE for agentic intelligence in the open”
It ranks as the #1 open text model on LMArena.
AirStreetCapital p.44/313 13%
2025 Air Street Capital The State of AI Report 2025
“OpenAI pivots from the “wrong side of history” to aligning with “America-first AI””
20k stars in weeks, but how much real usage?
20k · GitHub Stars
BondCap p.26/214 11%
2016 Bond Cap Internet Trends 2016
“Shanghai Area Over Past 2+ Decades = Illustrates Magnitude of China (& Emerging Asia) Growth”
AirStreetCapital p.28/213 12%
2024 Air Street Capital The State of AI Report 2024
“And could we distill transformers into hybrid models? It's...complicated.”
Mohawk can train Phi-Mamba and Hybrid-Phi-Mamba to achieve performance close to the teacher model.
3B · Average Accuracy (%)
AirStreetCapital p.32/213 14%
2024 Air Street Capital The State of AI Report 2024
“Web data is decanted openly at scale - proving quality is key”
FineWeb-edu, which filtered out samples scored below 3, outperformed FineWeb and all other open datasets, despite being significantly smaller.
15T ·
Deloitte p.25/190 12%
2025 The AI Dossier
Seeing is believing
AI can be used for style transferring, which allows consumers to see a digital rendering of clothes and other products on their own bodies, in their homes, and elsewhere.
AirStreetCapital p.20/188 10%
2021 Air Street Capital The State of AI Report 2021
“The ideas behind AlphaFold 2 rapidly diffused into academia and open source”
Half a year after DeepMind presented their AlphaFold 2 (AF2) method at the CASP14 conference, the Baker lab at the University of Washington created their own protein structure prediction system using
AirStreetCapital p.21/188 10%
2021 Air Street Capital The State of AI Report 2021
“Large language models can generate functional proteins that are unseen in nature”
AI-generated proteins across 5 families of antibacterial lysozymes show similar biological performance characteristics as their natural peers, even when their sequence similarity is only 44%.
44% · Percent functional
AirStreetCapital p.22/188 11%
2021 Air Street Capital The State of AI Report 2021
“Learning the language of Covid-19 to predict its evolution and escape mutants”
Language models trained on viral sequences can predict mutations that preserve infectivity but induce high antigenic change, akin to preserving “grammaticality” but inducing high “semantic change”.
AirStreetCapital p.23/188 11%
2021 Air Street Capital The State of AI Report 2021
“New state-of-the-art for predicting the 3D structure of RNA molecules”
In the RNA-Puzzles challenge, ARES selects the best Rosetta FARFAR2 model for each of four RNA molecules, beating humans and other methods, despite significant differences with its training set.
4.8 · RMSD (Å)
AirStreetCapital p.25/188 12%
2021 Air Street Capital The State of AI Report 2021
“Predicting and prioritising novel drug combinations, dosages, and timing for therapy”
At evaluation time, we can swap out the model's perturbation embedding to answer the counterfactual question "What would have the gene expression of this cell looked like, had it been treated differen
AirStreetCapital p.26/188 13%
2021 Air Street Capital The State of AI Report 2021
“Accelerating high-throughput virtual drug screening with model-guided search”
Deep learning models can learn drug-protein binding relationships from a small number of empirical experiments in order to help prioritise which areas of vast chemical spaces to virtually screen.
AirStreetCapital p.27/188 13%
2021 Air Street Capital The State of AI Report 2021
“Predicting chemical reaction performance using Transformers”
Reaction transformer encoder models fine-tuned on augmented reaction SMILES, outperform all previous approaches in predicting Buchwald-Hartwig reaction yields - an essential tool in the pharmaceutical
AirStreetCapital p.29/188 15%
2021 Air Street Capital The State of AI Report 2021
“Superhuman world models for Atari, but on a budget”
DreamerV2 is the first model-based RL agent trained on a single GPU to surpass human level performance on 55 popular tasks of the Atari benchmark.
AirStreetCapital p.21/177 11%
2020 Air Street Capital The State of AI Report 2020
“Low resource languages with limited training data are a beneficiary of large models”
Google made use of their large language models to deliver higher quality translations for languages with limited amounts of training data, for example Hansa and Uzbek.
AirStreetCapital p.23/177 12%
2020 Air Street Capital The State of AI Report 2020
“Yet, for some use cases like dialogue small, data-efficient models can trump large models”
Their model outperforms Google's BERT model in conversational applications, especially in low data regimes, suggesting BERT is far from a silver bullet for all NLP tasks.
68.2% · 1-vs-100 Accuracy
AirStreetCapital p.25/177 13%
2020 Air Street Capital The State of AI Report 2020
“Computer, please convert my code into another programming language”
An unsupervised machine translation model trained on GitHub projects with 1,000 parallel functions can translate 90% of these functions from C++ to Java and 57% of Python functions into C++ and succes
90% · translation success rate
AirStreetCapital p.26/177 14%
2020 Air Street Capital The State of AI Report 2020
“Computer, can you automatically repair my buggy programs too?”
Given a broken program and diagnostic feedback (compiler error message), DrRepair localizes an erroneous line and generates a repaired line.
Bain 2024 p.23/171 12%
SOUTHEAST ASIA’S GREEN ECONOMY 2024
“SEA renewed its commitments at COP28 and pushed for greater global support for transition”
Barclays 2024 p.16/145 10%
20240220 Barclays FY2023 Results and Investor Update Present
Acquisition of Tesco Bank Retail Banking
Expected completion H224 with initial annualised NII of c.£400m
£400m · NII
AirStreetCapital p.15/136 10%
2019 Air Street Capital The State of AI Report 2019
“What's next in RL: Play-driven learning for robots”
Despite not being trained on task-specific data, this system is capable of generalizing to 18 complex user-specified manipulation tasks with average success of 85.5%, outperforming individual models t
85.5% · Success rate
AirStreetCapital p.16/136 11%
2019 Air Street Capital The State of AI Report 2019
“What's next in RL: Learning dexterity using simulation and the world real”
New robotic learning platforms and sim-to-real enable impressive progress in manual dexterity.
AirStreetCapital p.18/136 12%
2019 Air Street Capital The State of AI Report 2019
“What’s next in RL: Learning dynamics models for online planning”
Within 500 episodes, PlaNet outperforms A3C trained from 100,000 episodes, on six simulator control tasks.
180x · Data efficiency gain
AirStreetCapital p.19/136 13%
2019 Air Street Capital The State of AI Report 2019
“What’s next in RL: Moving research into production environments”
Facebook release Horizon, the first open source end-to-end platform that uses applied RL to optimize systems in large-scale production environments, such as Messenger suggestions, video stream quality
AirStreetCapital p.20/136 14%
2019 Air Street Capital The State of AI Report 2019
“ML for life science: AlphaFold predicts de novo 3D structure of folded proteins”
Two deep CNNs work in concert to significantly outperform prior state-of-the-art, far earlier than expected.
AirStreetCapital p.13/114 10%
2022 Air Street Capital The State of AI Report 2022
“2021 Prediction: DeepMind’s breakthroughs in the physical sciences (2/3)”
DeepMind researchers used their framework in a collaboration with mathematics professors from the University of Sydney and the University of Oxford to (i) propose an algorithm that could solve a 40 ye
AirStreetCapital p.14/114 11%
2022 Air Street Capital The State of AI Report 2022
“2021 Prediction: DeepMind's breakthroughs in the physical sciences (3/3)”
DeepMind's approach not only helps speed up research in the field, but also boosts matrix multiplication based technology, that is AI, imaging, and essentially everything happening on our phones.
23.9% · Speed-up on Nvidia V100 GPU
AirStreetCapital p.15/114 12%
2022 Air Street Capital The State of AI Report 2022
“Reinforcement learning could be a core component of the next fusion breakthrough”
DeepMind trained a reinforcement learning system to adjust the magnetic coils of Lausanne's TCV (Variable Configuration tokamak).
AirStreetCapital p.16/114 13%
2022 Air Street Capital The State of AI Report 2022
“Predicting the structure of the entire known proteome: what could this unlock next?”
AlphaFold mentions in AI research literature is growing massively and is predicted to triple year on year.
358 · AI research literature mentions
AirStreetCapital p.17/114 14%
2022 Air Street Capital The State of AI Report 2022
“Language models for proteins: a familiar story of open source and scaled models”
Researchers independently applied language models to the problems of protein generation and structure prediction while scaling model parameter. They both report large benefits from scaling their model
6.37 · Validation Perplexity
KPMG 2024 p.15/106 13%
Our Impact Plan 2024
01 Governance
Through sustained innovation, we aim to consistently deliver superior quality to audited entities.
JPMorgan 2021 p.10/78 11%
jpmc esg report 2021
GREEN
By the end of 2030, we aim to finance and facilitate $1 trillion in support of the deployment of solutions for cleaner sources of energy and facilitating the transition to a low-carbon economy.
$106B · Green financing volume
JPMorgan 2021 p.11/78 13%
jpmc esg report 2021
DEVELOPMENT FINANCE
In 2021, approximately $117 billion of development finance activity - nearly 600 transactions - was qualified towards the Target.
$117B · Development Finance
MorganStanley 2022 p.9/77 10%
morgan stanley virtual hk summit march 2022
“Over 50 years of unbroken profitability”
Organic growth, Acquisitions, Regional expansion, Staff
50 years · Profitability
Barclays 2023 p.11/71 14%
Barclays Credit Bureau Forum 2023
“PowerCurve is recognised as an industry-leading Decisioning tool; faster deployment and integrated analytics unlocks new client opportunities”
Experian has been named a technology leader in Quadrant's 2023 SPARK Matrix: Digital Decisioning Platforms
35% · onboarding time reduction
BondCap p.10/67 13%
2011 Bond Cap Internet Trends 2011
“If it Plays in Peoria...Take it to Broadway! Especially in Mobile / Social World...”
100MM+ · Users
McKinsey 2020 p.9/65 12%
Chilean Hydrogen Pathway
“1.1 / Investment has been led by large industrials teaming up for large flagship projects – to reach technological and commercial maturity”
misc 2020 p.9/57 14%
COVID-19: Briefing Note
“South Korea: Rigorous investigation of outbreak clusters and rapidly scaled testing capabilities limited spread”
Localized outbreaks, including another infected church congregation, point to ongoing need for surveillance and response
Deloitte 2017 p.8/52 13%
Global Powers of Luxury Goods 2017 The new luxury consumer
“How does the need to operate in an omnichannel world affect operations?”
In conclusion, given ever-evolving customer preferences and increasing use of mobile platforms, the ability to switch seamlessly among different channels has become essential for personal luxury brand
misc 2022 p.8/49 14%
WHAT THE FUTURE: WELLNESS
Who is most optimistic about their health future?
How could we be intentional and purposeful about removing the barrier and driving better access to this affordable type care?
67% · Optimism (% Improve)
IPSOS 2022 p.8/49 14%
What The Future Wellness
Who is most optimistic about their health future?
How could we be intentional and purposeful about removing the barrier and driving better access to this affordable type care?
67% · Percentage of respondents expecting improvement
MorganStanley 2023 p.6/45 11%
20230608 172439 CWCU 9YRZMYZ26FO0PKXJ.1
“Projected +€250 million developments pipeline”
€250 million · Estimated development cost
IPSOS 2020 p.7/41 15%
2020 Effie UK Report in partnership with Ipsos MORI
Effie Case Study: The Bank of Antandec
The Bank of Antandec campaign exceeded all advertising category norms, increasing mortgage applications by 24% and growing market share by a whole percentage point, resulting in the best year for mort
24% · mortgage applications
misc 2020 p.7/41 15%
2020 Effie UK Report • In partnership withIpsos
Effie Case Study: The Bank of Antandec
The Bank of Antandec campaign exceeded all advertising category norms, increasing mortgage applications by 24% and growing market share by a whole percentage point, resulting in the best year for mort
24% · mortgage applications
IPSOS 2024 p.6/39 13%
ipsos the perils of perception 2024
Snapshots from our research programme
Read more about the research and explore the data on Ipsos' dedicated Perils of Perception site
22% · Public perception error rate
RolandBerger 2016 p.5/30 13%
Polish Digital Index
“Under patronage of Rzeczpospolita we have published a series of articles on the current state of digital readiness of polish industries”
Accenture 2021 p.4/23 13%
Blueprint for Service Success
“Building revenue streams from services”
An Industrial Internet of Things (IIoT) solution delivered new customer capabilities and services for its machinery aftermarket business.
Accenture 2023 p.4/23 13%
The evolving private equity playbook
Client example
Accenture has generated an estimated $40 million in savings for the client since 2013, representing an 89% internal rate of return per annum on an initial investment of $4.4 million to set up the serv
89% · internal rate of return
Accenture 2023 p.4/22 14%
2023 Post Parcel industry trends
Thought leadership
For more than 15 years, we have been following the Logistics & Post and Parcel industry and publishing leading Thought Leadership
Barclays 2022 p.4/22 14%
PR Barclays Presentation 9.06.22 FINAL Update
Colgate Value Creation History
Consistently among best results in Permian Basin.
20.2 · 6-mo Cumulative Boe/ft