SlideShare a Scribd company logo

Engineering tools for making smarter decisions .pptx

Tamir Dresher, profile picture
Tamir Dresher

The document outlines various engineering tools and methodologies for making informed decisions, focusing on resource allocation and project management strategies. It presents examples including optimal team compositions for project completion, communication method selection using multi-armed bandit strategies, and scaling SaaS products across regions with constraints. It emphasizes utilizing techniques like linear programming and Excel's built-in features to optimize decisions and efficiency in software engineering contexts.

Software
Related topics:
Decision-Making
1 of 56
Download to read offline
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
@tamir_dresher Engineering tools for making smarter decisions Tamir Dresher Director of Architecture @ Payoneer @tamir_dresher
@tamir_dresher Agenda 1. Home/Office/Freelancer 2. Choosing communication Under uncertainty 3. Resource allocation 4. Choosing the best library
@tamir_dresher Head of Architecture @ Who am I ? Software Engineering Lecturer Ruppin Academic Center & My Books:
@tamir_dresher @tamir_dresher Example 1 Work from home or office or take Freelancer
@tamir_dresher Example 1: Work from home or office or take Freelancer • We were given a project which we estimate will take 360 Man-Days to complete • 10 people team - FTEs (Full time employees) and Freelancers each with a different cost • FTEs – Can work at home or at the office – Office has a cost and if more than 2 days, cost increase • Freelancers – Has lesser productivity than FTE but lower cost – Cost increase if more than 8 freelancers • What is the ideal combination of FTEs and Freelancers, and how many office days?
@tamir_dresher Brute force =C6*(C10+C5*C8/5)+C4*IF(C4>8,C11,C9)+IF(C5>3,20,0)*C5
@tamir_dresher Tip 1 – Name you cells • Ctrl+F3 will open the Name Manager
@tamir_dresher Brute force
@tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster
@tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster Step 1 Write all values for variable 1 as rows, and all values for variable 2 as columns
@tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster Step 2 Reference you target function in the upper left cell of the table
@tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster Step 3 • Mark the table • Open What-If -> Data Table under the Data tab in the ribbon • Then reference the cell of the variable the top row represents and the variable the left column represents
@tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster
@tamir_dresher Tip 2 – Colorize your cells • Use Conditional Formatting to quickly find Minimal and Maximal values
@tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster
@tamir_dresher @tamir_dresher Example 2 Choosing the best communication method OR how to choose your next restaurant
@tamir_dresher Example 2: Choosing the best communication method • We are adding a new recommendation capability to our product and the team has 2 candidate possible algorithms. We are not sure which one will perform better (which one will lead to most clicks) • Naïve A/B Testing will expose each algorithm to 50% of the users for a period of time and then compare the results. • But If one algorithm is clearly better, we might be losing significant revenue while continuing to show the underperforming one during the test period
@tamir_dresher Multi Armed Bandit
@tamir_dresher Multi Armed Bandit balancing exploration (trying different options) and exploitation (sticking to the best-known option).
@tamir_dresher My restaurant search Corporate headquarters A B C D
@tamir_dresher My restaurant search Corporate headquarters A B C D which one should I pick from now on?
@tamir_dresher
@tamir_dresher My restaurant search Corporate headquarters A B C D Even a broken clock shows the right time twice a day
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 0 1
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 0 1 0.2 0.4 0.6 0.8 1 2 3 4 5
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 0 1 0.2 0.4 0.6 0.8 1 2 3 4 5 random.NextDouble(); -> 0.5
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability Random sample: 3 Random sample: 2 Random sample: 4 Random sample: 1
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability Random sample: 3 Random sample: 2 Random sample: 4 Random sample: 1
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability Random sample: 4 Random sample: 2 Random sample: 3 Random sample: 1
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
@tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
@tamir_dresher Multi Armed Bandit - summary • Balancing exploration (trying new options) and exploitation (choosing the best- known option) across multiple choices (e.g., restaurants, features). • The goal is to maximize reward over time by dynamically updating the selection process based on feedback. • Thompson sampling – generate a random value per each item/option by using it distribution array and select the item with the highest value – Beta distribution is a preferred way to do so
@tamir_dresher @tamir_dresher Example 3 SaaS expansion in the cloud and how to satisfy FinOps
@tamir_dresher Example 3: SaaS expansion in the cloud • Our SaaS product becomes massively popular and has been expanded into multiple regions in the cloud. Each region has different load and base latency. We want to optimize the cost while maintaining high performance and availability to all users • Decisions to make: – How many instance per each resource type in each region? – How to distribute the load between the regions? • Constraints: – Each region must provide at least 100,000 compute units (CU) – Latency must stay below a certain threshold – Some regions are critical and should have lower latency and availability – Our budget has a limit of 15K USD
@tamir_dresher Formalize the problem (1) Region Critical Base Load Base latency (ms) Routed load to (0..1) Instances US-East US-West EU- Central Asia- Pacific #Smal l #Medium #Larg e US-East 1 4500 50 ? ? ? ? ? ? ? US-West 0 3200 60 ? ? ? ? ? ? ? EU-Central 1 3250 58 ? ? ? ? ? ? ? Asia-Pacific 0 5500 80 ? ? ? ? ? ? ? 𝑅𝑒𝑔𝑖𝑜𝑛𝐿𝑎𝑡𝑒𝑛𝑐𝑦= 𝐵𝑎𝑠𝑒𝐿𝑎𝑡𝑒𝑛𝑐𝑦×𝑅𝑒𝑔𝑖𝑜𝑛𝐿𝑜𝑎𝑑 𝐶𝑜𝑚𝑝𝑢𝑡𝑒𝑈𝑛𝑖𝑡𝑠 × 1 𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙𝑖𝑡𝑦𝑊𝑒𝑖𝑔h𝑡 Increase Latency to virtually demand more resources Minimize Cost = (#Smallus-east+…+#SmallAsia)*Costsmall +…+ (#Largeus-east+…+#LargeAsia)*Costlarge
@tamir_dresher Formalize the problem (1) Region Critical Base Load Base latency (ms) Routed load to (0..1) Instances US-East US-West EU- Central Asia- Pacific #Smal l #Medium #Larg e US-East 1 4500 50 ? ? ? ? ? ? ? US-West 0 3200 60 ? ? ? ? ? ? ? EU-Central 1 3250 58 ? ? ? ? ? ? ? Asia-Pacific 0 5500 80 ? ? ? ? ? ? ? Constraints: • Regions routed load sum to 1 • Instance count are integers • Cost<15,000$ • Latency < MAX_LATENCY • Each region has minimal capacity Minimize Cost = (#Smallus-east+…+#SmallAsia)*Costsmall +…+ (#Largeus-east+…+#LargeAsia)*Costlarge
@tamir_dresher
@tamir_dresher Linear Programming/Optimization • Ideal for resource allocation problems
@tamir_dresher Linear Programming/Optimization • You can use Excel solver or use Google OR-Tools for .Net in your code
@tamir_dresher Linear Programming/Optimization • You can use Excel solver or use Google OR-Tools for .Net in your code https://github.com/tamirdresher/EngineeringToolsForSmarterDecisions/tree/master/LinearProgramming
@tamir_dresher Linear Programming - Real world scenario • How Uber Optimizes the Timing of Push Notifications using ML and Linear Programmi ng Where xi,t is a binary indicator of whether to send push i at time t, and si,t is the score (value) of sending push i at time t.
@tamir_dresher @tamir_dresher Example 4 Deciding which Framework/Library to use
@tamir_dresher Example 4: Deciding which Framework/Library to use • You're tasked with choosing a third-party library for the upcoming project. you need to consider several factors to ensure you pick the right one from 10 potential libraries. Each library is evaluated based on these attributes: 1. Ease of Use 2. Performance 3. Cost 4. Support and Community 5. Documentation Quality 6. Downloads/Stars
@tamir_dresher Example 4: Deciding which Framework/Library to use Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentati on Quality (1-10) Downloads Library1 5 700 500 0 7 1000 Library2 7 650 1200 1 6 2000 Library3 8 699 450 0 9 500 Library4 5 520 600 1 5 900 Library5 6 200 150 1 8 250 Library6 9 600 800 0 4 740 Library7 4 320 350 1 2 800 Library8 5 875 300 1 8 600 Library9 2 900 450 1 2 700 Library1 0 10 1000 10 1 10 1 Issues: • Scale is not comparable • For cost, larger value should give lower score • How should we take #Downloads into account
@tamir_dresher Normalizing the data • Each attribute has different scales of importance attributes like cost (which have a wider range) have a disproportionate impact on the decision • min-max normalization – transform to a range of [0,1] • Z-score – transforms to a range of [-1,1] min-max makes numbers fit between 0 and 1, while z-score shows how far each number is from the average, using the idea of "how spread out" the numbers are.
@tamir_dresher Normalized attributes Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentation Quality (1-10) Downloa ds Library 1 -0.478 0.224 0.060 -1.528 0.338 0.497 Library 2 0.391 0.015 2.255 0.655 -0.038 2.478 Library 3 0.826 0.220 -0.097 -1.528 1.089 -0.493 Library 4 -0.478 -0.529 0.373 0.655 -0.413 0.299 Library 5 -0.043 -1.868 -1.038 0.655 0.714 -0.989 Library 6 1.261 -0.194 1.001 -1.528 -0.789 -0.018 =(B2-AVERAGE(B:B))/STDEV.P(B:B) Cost should order should be reversed. So we will multiply by -1
@tamir_dresher Normalized attributes with cost order fix Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentation Quality (1-10) Downloa ds Library 1 -0.478 0.224 -0.060 -1.528 0.338 0.497 Library 2 0.391 0.015 -2.255 0.655 -0.038 2.478 Library 3 0.826 0.220 0.097 -1.528 1.089 -0.493 Library 4 -0.478 -0.529 -0.373 0.655 -0.413 0.299 Library 5 -0.043 -1.868 1.038 0.655 0.714 -0.989 Library 6 1.261 -0.194 -1.001 -1.528 -0.789 -0.018
@tamir_dresher Normalized attributes with cost order fix – with avg score Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentation Quality (1-10) Downloa ds Library 1 -0.478 0.224 -0.060 -1.528 0.338 0.497 Library 2 0.391 0.015 -2.255 0.655 -0.038 2.478 Library 3 0.826 0.220 0.097 -1.528 1.089 -0.493 Library 4 -0.478 -0.529 -0.373 0.655 -0.413 0.299 Library 5 -0.043 -1.868 1.038 0.655 0.714 -0.989 Library 6 1.261 -0.194 -1.001 -1.528 -0.789 -0.018 Score -0.16769 0.20769 0.035262 -0.13999 -0.08225 -0.37801 -0.44203 0.353136 -0.26778 0.881655
@tamir_dresher
@tamir_dresher IMDB Score 𝐼𝑀𝐷𝐵𝑆𝑐𝑜𝑟𝑒= 𝑁𝑢𝑚𝑏𝑒𝑟𝑂𝑓𝑉𝑜𝑡𝑒𝑠∗𝑀𝑜𝑣𝑖𝑒𝐴𝑣𝑔𝑅𝑎𝑡𝑖𝑛𝑔+𝑚𝑖𝑛𝑉𝑜𝑡𝑒𝑠∗ 𝐴𝑣𝑔𝑅𝑎𝑡𝑖𝑛𝑔𝑂𝑓𝐴𝑙𝑙𝑀𝑜𝑣𝑖𝑒𝑠 𝑁𝑢𝑚𝑏𝑒𝑟𝑂𝑓𝑉𝑜𝑡𝑒𝑠+𝑚𝑖𝑛𝑉𝑜𝑡𝑒𝑠 𝐿𝑖𝑏𝑟𝑎𝑟𝑦 𝑆𝑐𝑜𝑟𝑒= 𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠∗𝐿𝑖𝑏𝐴𝑣𝑔𝑆𝑐𝑜𝑟𝑒+𝑚𝑖𝑛𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠∗ 𝐴𝑣𝑔𝑆𝑐𝑜𝑟𝑒 𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠+𝑚𝑖𝑛𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠
@tamir_dresher Example 4: Deciding which Framework/Library to use Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentati on Quality (1-10) Downloads Score (Normalize d + IMDB) Library 1 5 700 500 0 7 1000 -0.27329 Library 2 7 650 1200 1 6 2000 -0.23461 Library 3 8 699 450 0 9 500 0.1175 Library 4 5 520 600 1 5 900 -0.20499 Library 5 6 200 150 1 8 250 0.070731 Library Final results
@tamir_dresher Summary 1. Home/Office/Freelancer – Brute force & What-If 2. Choosing communication Under uncertainty – Multi Armed Bandit 3. Resource allocation Linear Optimization 4. Choosing the best library – IMDB Score https://github.com/tamirdresher/EngineeringToolsForSmarterDecisio ns

More Related Content

PDF
Scott Clark, Software Engineer, Yelp at MLconf SF
MLconf
 
PDF
Optimal Learning for Fun and Profit with MOE
Yelp Engineering
 
PDF
heuristic search Techniques and game playing.pdf
vijeta3feb
 
PDF
Introduction to search and optimisation for the design theorist
Akin Osman Kazakci
 
PDF
"Optimal Learning for Fun and Profit" by Scott Clark (Presented at The Yelp E...
Yelp Engineering
 
PPTX
Applied Data Science for monetization: pitfalls, common misconceptions, and n...
DevGAMM Conference
 
PPTX
ETCS262A-Analysis of design Algorithm.pptx
RahulSingh190790
 
PPTX
I love the smell of data in the morning (getting started with data science) ...
Troy Magennis
 
Scott Clark, Software Engineer, Yelp at MLconf SF
MLconf
 
Optimal Learning for Fun and Profit with MOE
Yelp Engineering
 
heuristic search Techniques and game playing.pdf
vijeta3feb
 
Introduction to search and optimisation for the design theorist
Akin Osman Kazakci
 
"Optimal Learning for Fun and Profit" by Scott Clark (Presented at The Yelp E...
Yelp Engineering
 
Applied Data Science for monetization: pitfalls, common misconceptions, and n...
DevGAMM Conference
 
ETCS262A-Analysis of design Algorithm.pptx
RahulSingh190790
 
I love the smell of data in the morning (getting started with data science) ...
Troy Magennis
 

Similar to Engineering tools for making smarter decisions .pptx (20)

PDF
Backtracking & branch and bound
Vipul Chauhan
 
PDF
FADML 10 PPC Solving NP-Hard Problems.pdf
Yelah1
 
PDF
Monte Carlo Simulation for project estimates v1.0
PMILebanonChapter
 
PPT
Elak3 need of greedy for design and analysis of algorithms.ppt
Elakkiya Rajasekar
 
PPT
Software Testing-Dynamic testing technique-Mazenet solution
Mazenetsolution
 
PDF
My Favorite Calc Code
Alithya
 
PPTX
Artificial Intelligence Alpha Beta pruning.pptx
Kalpana Mohan
 
PDF
Eficiency and Low Cost: Pro Tips for you to save 50% of your money with Googl...
Daniel Cukier
 
PPTX
FAI UNIT 6FAI UNIT 7 all important questions to be noted max pass
22eg105n11
 
PDF
Problem Solving Techniques For Evolutionary Design
Naresh Jain
 
PPTX
Algorithm and data science foundation ISC
MarufFarhanRigan1
 
PDF
LKCE16 - Estimation made easy by Pawel Brodzinski and Tomek Rusilko
Lean Kanban Central Europe
 
PDF
DM 3.pdf
AhmedEldeeb984938
 
PDF
Data Science Toolkit for Product Managers
Mahmoud Jalajel
 
PDF
Data science toolkit for product managers
ProductFolks
 
PDF
Bandit algorithms for website optimization - A summary
Kwanghee Choi
 
PPT
Parallel_Algorithms_In_Combinatorial_Optimization_Problems.ppt
BinayakMukherjee4
 
PPT
126622gghyytgggffggggggfsssssssssdff70.ppt
ShahidHussain66691
 
PPTX
Risk Management and Reliable Forecasting using Un-reliable Data (magennis) - ...
Troy Magennis
 
PDF
"Computational Support for Functionality Selection in Interaction Design" CHI...
Aalto University
 
Backtracking & branch and bound
Vipul Chauhan
 
FADML 10 PPC Solving NP-Hard Problems.pdf
Yelah1
 
Monte Carlo Simulation for project estimates v1.0
PMILebanonChapter
 
Elak3 need of greedy for design and analysis of algorithms.ppt
Elakkiya Rajasekar
 
Software Testing-Dynamic testing technique-Mazenet solution
Mazenetsolution
 
My Favorite Calc Code
Alithya
 
Artificial Intelligence Alpha Beta pruning.pptx
Kalpana Mohan
 
Eficiency and Low Cost: Pro Tips for you to save 50% of your money with Googl...
Daniel Cukier
 
FAI UNIT 6FAI UNIT 7 all important questions to be noted max pass
22eg105n11
 
Problem Solving Techniques For Evolutionary Design
Naresh Jain
 
Algorithm and data science foundation ISC
MarufFarhanRigan1
 
LKCE16 - Estimation made easy by Pawel Brodzinski and Tomek Rusilko
Lean Kanban Central Europe
 
DM 3.pdf
AhmedEldeeb984938
 
Data Science Toolkit for Product Managers
Mahmoud Jalajel
 
Data science toolkit for product managers
ProductFolks
 
Bandit algorithms for website optimization - A summary
Kwanghee Choi
 
Parallel_Algorithms_In_Combinatorial_Optimization_Problems.ppt
BinayakMukherjee4
 
126622gghyytgggffggggggfsssssssssdff70.ppt
ShahidHussain66691
 
Risk Management and Reliable Forecasting using Un-reliable Data (magennis) - ...
Troy Magennis
 
"Computational Support for Functionality Selection in Interaction Design" CHI...
Aalto University
 
Ad

More from Tamir Dresher (20)

PDF
NET Aspire - NET Conf IL 2024 - Tamir Dresher.pdf
Tamir Dresher
 
PPTX
Tamir Dresher - DotNet 7 What's new.pptx
Tamir Dresher
 
PPTX
Tamir Dresher - What’s new in ASP.NET Core 6
Tamir Dresher
 
PPTX
Tamir Dresher - Async Streams in C#
Tamir Dresher
 
PPTX
Anatomy of a data driven architecture - Tamir Dresher
Tamir Dresher
 
PPTX
Tamir Dresher Clarizen adventures with the wild GC during the holiday season
Tamir Dresher
 
PDF
Debugging tricks you wish you knew Tamir Dresher - Odessa 2019
Tamir Dresher
 
PDF
From zero to hero with the actor model - Tamir Dresher - Odessa 2019
Tamir Dresher
 
PPTX
Tamir Dresher - Demystifying the Core of .NET Core
Tamir Dresher
 
PPTX
Breaking the monolith to microservice with Docker and Kubernetes (k8s)
Tamir Dresher
 
PPTX
.Net december 2017 updates - Tamir Dresher
Tamir Dresher
 
PPTX
Testing time and concurrency Rx
Tamir Dresher
 
PPTX
Building responsive application with Rx - confoo - tamir dresher
Tamir Dresher
 
PPTX
.NET Debugging tricks you wish you knew tamir dresher
Tamir Dresher
 
PPTX
From Zero to the Actor Model (With Akka.Net) - CodeMash2017 - Tamir Dresher
Tamir Dresher
 
PPTX
Building responsive applications with Rx - CodeMash2017 - Tamir Dresher
Tamir Dresher
 
PPTX
Debugging tricks you wish you knew - Tamir Dresher
Tamir Dresher
 
PPTX
Rx 101 - Tamir Dresher - Copenhagen .NET User Group
Tamir Dresher
 
PPTX
Cloud patterns - NDC Oslo 2016 - Tamir Dresher
Tamir Dresher
 
PPTX
Reactiveness All The Way - SW Architecture 2015 Conference
Tamir Dresher
 
NET Aspire - NET Conf IL 2024 - Tamir Dresher.pdf
Tamir Dresher
 
Tamir Dresher - DotNet 7 What's new.pptx
Tamir Dresher
 
Tamir Dresher - What’s new in ASP.NET Core 6
Tamir Dresher
 
Tamir Dresher - Async Streams in C#
Tamir Dresher
 
Anatomy of a data driven architecture - Tamir Dresher
Tamir Dresher
 
Tamir Dresher Clarizen adventures with the wild GC during the holiday season
Tamir Dresher
 
Debugging tricks you wish you knew Tamir Dresher - Odessa 2019
Tamir Dresher
 
From zero to hero with the actor model - Tamir Dresher - Odessa 2019
Tamir Dresher
 
Tamir Dresher - Demystifying the Core of .NET Core
Tamir Dresher
 
Breaking the monolith to microservice with Docker and Kubernetes (k8s)
Tamir Dresher
 
.Net december 2017 updates - Tamir Dresher
Tamir Dresher
 
Testing time and concurrency Rx
Tamir Dresher
 
Building responsive application with Rx - confoo - tamir dresher
Tamir Dresher
 
.NET Debugging tricks you wish you knew tamir dresher
Tamir Dresher
 
From Zero to the Actor Model (With Akka.Net) - CodeMash2017 - Tamir Dresher
Tamir Dresher
 
Building responsive applications with Rx - CodeMash2017 - Tamir Dresher
Tamir Dresher
 
Debugging tricks you wish you knew - Tamir Dresher
Tamir Dresher
 
Rx 101 - Tamir Dresher - Copenhagen .NET User Group
Tamir Dresher
 
Cloud patterns - NDC Oslo 2016 - Tamir Dresher
Tamir Dresher
 
Reactiveness All The Way - SW Architecture 2015 Conference
Tamir Dresher
 
Ad

Recently uploaded (20)

PPTX
Mastering-Cybersecurity-The-Crucial-Role-of-Antivirus-Support-Services.pptx
akkie9682
 
PDF
top salesforce developer skills in 2025.pdf
CloudMetic
 
PDF
Why TechBuilder is the Future of Pickup and Delivery App Development (1).pdf
TechBuilder
 
PPTX
Materi-Enum-and-Record-Data-Type (1).pptx
RanuFajar1
 
PDF
Claude Code: Everyone is a 10x Developer - A Comprehensive AI-Powered CLI Tool
William Chong
 
PDF
Build Multi-agent using Agent Development Kit
FadyIbrahim23
 
PPTX
AIRLINE PRICE API | FLIGHT API COST |
philipnathen82
 
PDF
Addressing The Cult of Project Management Tools-Why Disconnected Work is Hold...
OnePlan Solutions
 
PDF
PTS Company Brochure 2025 (1).pdf.......
pts464036
 
PPTX
Transform Your Business with a Software ERP System
Canada Software Company
 
PPTX
Agentic AI : A Practical Guide. Undersating, Implementing and Scaling Autono...
IT IDOL Technologies
 
PPTX
FLIGHT TICKET RESERVATION SYSTEM | FLIGHT BOOKING ENGINE API
philipnathen82
 
PDF
Digital Strategies for Manufacturing Companies
Virendra Rai, PMP
 
PDF
Softaken Excel to vCard Converter Software.pdf
markwillsonmw004
 
PDF
How to Confidently Manage Project Budgets
OnePlan Solutions
 
PPTX
VVF-Customer-Presentation2025-Ver1.9.pptx
blackmambaettijean
 
PDF
How Creative Agencies Leverage Project Management Software.pdf
Orangescrum
 
PDF
medical staffing services at VALiNTRY
Tiyan Grayson
 
PPTX
Odoo POS Development Services by CandidRoot Solutions
CandidRoot Solutions Private Limited
 
PDF
Understanding Forklifts - TECH EHS Solution
TECH EHS Solution
 
Mastering-Cybersecurity-The-Crucial-Role-of-Antivirus-Support-Services.pptx
akkie9682
 
top salesforce developer skills in 2025.pdf
CloudMetic
 
Why TechBuilder is the Future of Pickup and Delivery App Development (1).pdf
TechBuilder
 
Materi-Enum-and-Record-Data-Type (1).pptx
RanuFajar1
 
Claude Code: Everyone is a 10x Developer - A Comprehensive AI-Powered CLI Tool
William Chong
 
Build Multi-agent using Agent Development Kit
FadyIbrahim23
 
AIRLINE PRICE API | FLIGHT API COST |
philipnathen82
 
Addressing The Cult of Project Management Tools-Why Disconnected Work is Hold...
OnePlan Solutions
 
PTS Company Brochure 2025 (1).pdf.......
pts464036
 
Transform Your Business with a Software ERP System
Canada Software Company
 
Agentic AI : A Practical Guide. Undersating, Implementing and Scaling Autono...
IT IDOL Technologies
 
FLIGHT TICKET RESERVATION SYSTEM | FLIGHT BOOKING ENGINE API
philipnathen82
 
Digital Strategies for Manufacturing Companies
Virendra Rai, PMP
 
Softaken Excel to vCard Converter Software.pdf
markwillsonmw004
 
How to Confidently Manage Project Budgets
OnePlan Solutions
 
VVF-Customer-Presentation2025-Ver1.9.pptx
blackmambaettijean
 
How Creative Agencies Leverage Project Management Software.pdf
Orangescrum
 
medical staffing services at VALiNTRY
Tiyan Grayson
 
Odoo POS Development Services by CandidRoot Solutions
CandidRoot Solutions Private Limited
 
Understanding Forklifts - TECH EHS Solution
TECH EHS Solution
 

Engineering tools for making smarter decisions .pptx

  • 1. @tamir_dresher Engineering tools for making smarter decisions Tamir Dresher Director of Architecture @ Payoneer @tamir_dresher
  • 2. @tamir_dresher Agenda 1. Home/Office/Freelancer 2. Choosing communication Under uncertainty 3. Resource allocation 4. Choosing the best library
  • 3. @tamir_dresher Head of Architecture @ Who am I ? Software Engineering Lecturer Ruppin Academic Center & My Books:
  • 4. @tamir_dresher @tamir_dresher Example 1 Work from home or office or take Freelancer
  • 5. @tamir_dresher Example 1: Work from home or office or take Freelancer • We were given a project which we estimate will take 360 Man-Days to complete • 10 people team - FTEs (Full time employees) and Freelancers each with a different cost • FTEs – Can work at home or at the office – Office has a cost and if more than 2 days, cost increase • Freelancers – Has lesser productivity than FTE but lower cost – Cost increase if more than 8 freelancers • What is the ideal combination of FTEs and Freelancers, and how many office days?
  • 7. @tamir_dresher Tip 1 – Name you cells • Ctrl+F3 will open the Name Manager
  • 9. @tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster
  • 10. @tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster Step 1 Write all values for variable 1 as rows, and all values for variable 2 as columns
  • 11. @tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster Step 2 Reference you target function in the upper left cell of the table
  • 12. @tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster Step 3 • Mark the table • Open What-If -> Data Table under the Data tab in the ribbon • Then reference the cell of the variable the top row represents and the variable the left column represents
  • 13. @tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster
  • 14. @tamir_dresher Tip 2 – Colorize your cells • Use Conditional Formatting to quickly find Minimal and Maximal values
  • 15. @tamir_dresher Brute force • Let’s run over all the possible values of our variables • We can run a script BUT Excel’s build-in What-If feature is much faster
  • 16. @tamir_dresher @tamir_dresher Example 2 Choosing the best communication method OR how to choose your next restaurant
  • 17. @tamir_dresher Example 2: Choosing the best communication method • We are adding a new recommendation capability to our product and the team has 2 candidate possible algorithms. We are not sure which one will perform better (which one will lead to most clicks) • Naïve A/B Testing will expose each algorithm to 50% of the users for a period of time and then compare the results. • But If one algorithm is clearly better, we might be losing significant revenue while continuing to show the underperforming one during the test period
  • 19. @tamir_dresher Multi Armed Bandit balancing exploration (trying different options) and exploitation (sticking to the best-known option).
  • 21. @tamir_dresher My restaurant search Corporate headquarters A B C D which one should I pick from now on?
  • 23. @tamir_dresher My restaurant search Corporate headquarters A B C D Even a broken clock shows the right time twice a day
  • 24. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability
  • 25. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 0 1
  • 26. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 0 1 0.2 0.4 0.6 0.8 1 2 3 4 5
  • 27. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 0 1 0.2 0.4 0.6 0.8 1 2 3 4 5 random.NextDouble(); -> 0.5
  • 28. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability Random sample: 3 Random sample: 2 Random sample: 4 Random sample: 1
  • 29. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability Random sample: 3 Random sample: 2 Random sample: 4 Random sample: 1
  • 30. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability
  • 31. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
  • 32. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
  • 33. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability Random sample: 4 Random sample: 2 Random sample: 3 Random sample: 1
  • 34. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
  • 35. @tamir_dresher My restaurant search Corporate headquarters A B C D Multi Armed Bandit approach (Thompson sampling) 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 probability 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 probability
  • 36. @tamir_dresher Multi Armed Bandit - summary • Balancing exploration (trying new options) and exploitation (choosing the best- known option) across multiple choices (e.g., restaurants, features). • The goal is to maximize reward over time by dynamically updating the selection process based on feedback. • Thompson sampling – generate a random value per each item/option by using it distribution array and select the item with the highest value – Beta distribution is a preferred way to do so
  • 37. @tamir_dresher @tamir_dresher Example 3 SaaS expansion in the cloud and how to satisfy FinOps
  • 38. @tamir_dresher Example 3: SaaS expansion in the cloud • Our SaaS product becomes massively popular and has been expanded into multiple regions in the cloud. Each region has different load and base latency. We want to optimize the cost while maintaining high performance and availability to all users • Decisions to make: – How many instance per each resource type in each region? – How to distribute the load between the regions? • Constraints: – Each region must provide at least 100,000 compute units (CU) – Latency must stay below a certain threshold – Some regions are critical and should have lower latency and availability – Our budget has a limit of 15K USD
  • 39. @tamir_dresher Formalize the problem (1) Region Critical Base Load Base latency (ms) Routed load to (0..1) Instances US-East US-West EU- Central Asia- Pacific #Smal l #Medium #Larg e US-East 1 4500 50 ? ? ? ? ? ? ? US-West 0 3200 60 ? ? ? ? ? ? ? EU-Central 1 3250 58 ? ? ? ? ? ? ? Asia-Pacific 0 5500 80 ? ? ? ? ? ? ? 𝑅𝑒𝑔𝑖𝑜𝑛𝐿𝑎𝑡𝑒𝑛𝑐𝑦= 𝐵𝑎𝑠𝑒𝐿𝑎𝑡𝑒𝑛𝑐𝑦×𝑅𝑒𝑔𝑖𝑜𝑛𝐿𝑜𝑎𝑑 𝐶𝑜𝑚𝑝𝑢𝑡𝑒𝑈𝑛𝑖𝑡𝑠 × 1 𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙𝑖𝑡𝑦𝑊𝑒𝑖𝑔h𝑡 Increase Latency to virtually demand more resources Minimize Cost = (#Smallus-east+…+#SmallAsia)*Costsmall +…+ (#Largeus-east+…+#LargeAsia)*Costlarge
  • 40. @tamir_dresher Formalize the problem (1) Region Critical Base Load Base latency (ms) Routed load to (0..1) Instances US-East US-West EU- Central Asia- Pacific #Smal l #Medium #Larg e US-East 1 4500 50 ? ? ? ? ? ? ? US-West 0 3200 60 ? ? ? ? ? ? ? EU-Central 1 3250 58 ? ? ? ? ? ? ? Asia-Pacific 0 5500 80 ? ? ? ? ? ? ? Constraints: • Regions routed load sum to 1 • Instance count are integers • Cost<15,000$ • Latency < MAX_LATENCY • Each region has minimal capacity Minimize Cost = (#Smallus-east+…+#SmallAsia)*Costsmall +…+ (#Largeus-east+…+#LargeAsia)*Costlarge
  • 43. @tamir_dresher Linear Programming/Optimization • You can use Excel solver or use Google OR-Tools for .Net in your code
  • 44. @tamir_dresher Linear Programming/Optimization • You can use Excel solver or use Google OR-Tools for .Net in your code https://github.com/tamirdresher/EngineeringToolsForSmarterDecisions/tree/master/LinearProgramming
  • 45. @tamir_dresher Linear Programming - Real world scenario • How Uber Optimizes the Timing of Push Notifications using ML and Linear Programmi ng Where xi,t is a binary indicator of whether to send push i at time t, and si,t is the score (value) of sending push i at time t.
  • 47. @tamir_dresher Example 4: Deciding which Framework/Library to use • You're tasked with choosing a third-party library for the upcoming project. you need to consider several factors to ensure you pick the right one from 10 potential libraries. Each library is evaluated based on these attributes: 1. Ease of Use 2. Performance 3. Cost 4. Support and Community 5. Documentation Quality 6. Downloads/Stars
  • 48. @tamir_dresher Example 4: Deciding which Framework/Library to use Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentati on Quality (1-10) Downloads Library1 5 700 500 0 7 1000 Library2 7 650 1200 1 6 2000 Library3 8 699 450 0 9 500 Library4 5 520 600 1 5 900 Library5 6 200 150 1 8 250 Library6 9 600 800 0 4 740 Library7 4 320 350 1 2 800 Library8 5 875 300 1 8 600 Library9 2 900 450 1 2 700 Library1 0 10 1000 10 1 10 1 Issues: • Scale is not comparable • For cost, larger value should give lower score • How should we take #Downloads into account
  • 49. @tamir_dresher Normalizing the data • Each attribute has different scales of importance attributes like cost (which have a wider range) have a disproportionate impact on the decision • min-max normalization – transform to a range of [0,1] • Z-score – transforms to a range of [-1,1] min-max makes numbers fit between 0 and 1, while z-score shows how far each number is from the average, using the idea of "how spread out" the numbers are.
  • 50. @tamir_dresher Normalized attributes Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentation Quality (1-10) Downloa ds Library 1 -0.478 0.224 0.060 -1.528 0.338 0.497 Library 2 0.391 0.015 2.255 0.655 -0.038 2.478 Library 3 0.826 0.220 -0.097 -1.528 1.089 -0.493 Library 4 -0.478 -0.529 0.373 0.655 -0.413 0.299 Library 5 -0.043 -1.868 -1.038 0.655 0.714 -0.989 Library 6 1.261 -0.194 1.001 -1.528 -0.789 -0.018 =(B2-AVERAGE(B:B))/STDEV.P(B:B) Cost should order should be reversed. So we will multiply by -1
  • 51. @tamir_dresher Normalized attributes with cost order fix Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentation Quality (1-10) Downloa ds Library 1 -0.478 0.224 -0.060 -1.528 0.338 0.497 Library 2 0.391 0.015 -2.255 0.655 -0.038 2.478 Library 3 0.826 0.220 0.097 -1.528 1.089 -0.493 Library 4 -0.478 -0.529 -0.373 0.655 -0.413 0.299 Library 5 -0.043 -1.868 1.038 0.655 0.714 -0.989 Library 6 1.261 -0.194 -1.001 -1.528 -0.789 -0.018
  • 52. @tamir_dresher Normalized attributes with cost order fix – with avg score Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentation Quality (1-10) Downloa ds Library 1 -0.478 0.224 -0.060 -1.528 0.338 0.497 Library 2 0.391 0.015 -2.255 0.655 -0.038 2.478 Library 3 0.826 0.220 0.097 -1.528 1.089 -0.493 Library 4 -0.478 -0.529 -0.373 0.655 -0.413 0.299 Library 5 -0.043 -1.868 1.038 0.655 0.714 -0.989 Library 6 1.261 -0.194 -1.001 -1.528 -0.789 -0.018 Score -0.16769 0.20769 0.035262 -0.13999 -0.08225 -0.37801 -0.44203 0.353136 -0.26778 0.881655
  • 54. @tamir_dresher IMDB Score 𝐼𝑀𝐷𝐵𝑆𝑐𝑜𝑟𝑒= 𝑁𝑢𝑚𝑏𝑒𝑟𝑂𝑓𝑉𝑜𝑡𝑒𝑠∗𝑀𝑜𝑣𝑖𝑒𝐴𝑣𝑔𝑅𝑎𝑡𝑖𝑛𝑔+𝑚𝑖𝑛𝑉𝑜𝑡𝑒𝑠∗ 𝐴𝑣𝑔𝑅𝑎𝑡𝑖𝑛𝑔𝑂𝑓𝐴𝑙𝑙𝑀𝑜𝑣𝑖𝑒𝑠 𝑁𝑢𝑚𝑏𝑒𝑟𝑂𝑓𝑉𝑜𝑡𝑒𝑠+𝑚𝑖𝑛𝑉𝑜𝑡𝑒𝑠 𝐿𝑖𝑏𝑟𝑎𝑟𝑦 𝑆𝑐𝑜𝑟𝑒= 𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠∗𝐿𝑖𝑏𝐴𝑣𝑔𝑆𝑐𝑜𝑟𝑒+𝑚𝑖𝑛𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠∗ 𝐴𝑣𝑔𝑆𝑐𝑜𝑟𝑒 𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠+𝑚𝑖𝑛𝐷𝑜𝑤𝑛𝑙𝑜𝑎𝑑𝑠
  • 55. @tamir_dresher Example 4: Deciding which Framework/Library to use Library Ease of use (1- 10) Performance (TPS) Cost($ ) Support (1,0) Documentati on Quality (1-10) Downloads Score (Normalize d + IMDB) Library 1 5 700 500 0 7 1000 -0.27329 Library 2 7 650 1200 1 6 2000 -0.23461 Library 3 8 699 450 0 9 500 0.1175 Library 4 5 520 600 1 5 900 -0.20499 Library 5 6 200 150 1 8 250 0.070731 Library Final results
  • 56. @tamir_dresher Summary 1. Home/Office/Freelancer – Brute force & What-If 2. Choosing communication Under uncertainty – Multi Armed Bandit 3. Resource allocation Linear Optimization 4. Choosing the best library – IMDB Score https://github.com/tamirdresher/EngineeringToolsForSmarterDecisio ns

Editor's Notes

  • #18: Bandit - a thief or someone who belongs to a gang. A slot machine is like a thief of your time and resources
  • #19: Bandit - a thief or someone who belongs to a gang. A slot machine is like a thief of your time and resources
  • #24: Assign distribution array for each restaurant – uniform at first draw a number from each one
  • #25: Assign distribution array for each restaurant – uniform at first draw a number from each one
  • #26: Assign distribution array for each restaurant – uniform at first draw a number from each one
  • #27: Assign distribution array for each restaurant – uniform at first draw a number from each one
  • #28: Assign distribution array for each restaurant – uniform at first draw a number from each one
  • #29: Assign distribution array for each restaurant – uniform at first draw a number from each one
  • #30: Assign distribution array for each restaurant – uniform at first
  • #31: Assign distribution array for each restaurant – uniform at first
  • #32: Assign distribution array for each restaurant – uniform at first
  • #33: Assign distribution array for each restaurant – uniform at first
  • #34: Assign distribution array for each restaurant – uniform at first
  • #35: Assign distribution array for each restaurant – uniform at first