Introduction to augmented engineering

Published on Slideshow
Static slideshow
Download PDF version
Download PDF version
Embed video
Share video
Ask about this video

Scene 1 (0s)

[image] A rainbow colored object with black background Description automatically generated.

Scene 2 (9s)

[Audio] Intro presenters So now lets welcome Andy Vickers..

Scene 3 (34s)

Capgemini announcements on Generative AI. I am convinced that Generative AI will play a major role in the digital transition. The Group will invest €2 billion in Artificial Intelligence to build its leadership in this breakthrough technology, that must be deployed responsibly, reliably, and sustainably. We are developing a portfolio of industry-specific offers and signing strategic partnerships..

Scene 4 (1m 23s)

[image] Capture d expositions longues d un train en mouvement.

Scene 5 (1m 33s)

Introduction to augmented engineering. Elevate your Possible.

Scene 6 (1m 50s)

Context Regulated and standardised environment Safety, security, reliability, … Trust a foundational topic Correctness & precision are fundamental Multi-modal – more than words Text, engineering diagrams, models….

Scene 7 (2m 50s)

A red cube with many balls Description automatically generated.

Scene 8 (3m 21s)

Contextual. Statistical. Symbolic. Probabilistic.

Scene 9 (3m 49s)

General High-Level AI Lifecycle. Database with solid fill.

Scene 10 (4m 47s)

1. Training The machine learning model is shown examples of the desired data, and a model is built which encodes the relationships, patterns and sequences within it..

Scene 11 (5m 46s)

Retrieval Augmented Generation (RAG). Generative AI.

Scene 12 (6m 29s)

Issues. Fine-tuning. Pre-trained model. Database with solid fill.

Scene 13 (7m 12s)

Combine language models with other types of model and technology to develop reasoning and rationality.

Scene 14 (8m 13s)

ER&D GenAI Policy. [image] GENERA A1- MANDAT CUIDELINES.

Scene 15 (8m 28s)

REFERENTIAL ARCHITECTURE. Our referential architecture describes a standard approach for developing Hybrid AI solutions in Capgemini Engineering. This means we : Don’t need to reinvent the wheel for each solution. Can use standard terminology in a consistent way when communicating with our colleagues and customers..

Scene 16 (8m 56s)

[image] Brainstorming de l quipe commerciale. HYBRID AI SOLUTIONEERS ABLE TO HELP BUILD END-TO-END AI/GENAI ARCHITECTURE.

Scene 17 (9m 5s)

[image] A group of orange objects on a white plate Description automatically generated.

Scene 18 (9m 36s)

[Audio] Per offer: Scope pre-requisite pilot client status of offer development.

Scene 19 (10m 14s)

[Audio] Per offer: Scope pre-requisite pilot client status of offer development.

Scene 20 (10m 41s)

Results: 12 Arrow Electronics Seamless Integration & Transformation for CPQ & Reselling Platform Learn more -5 GSK GenAI — GenAI R&D Assistant for New Drug Discovery and Faster Identification of New Candidate Molecule Learn more Koch Industries GenAl assistant for first-draft Request for Quote/Proposal to best answer customer needs Learn more AstraZeneca GenAI — GenAI R&D Assistant to Predict Adverse Drug Reactions With 75% Accuracy Prior to Market Release Learn more Voda fone GenAl — Experiment GenAl Technologies to Accelerate the Development of Python Code Learn more -5 GenAI — GenAI Synthetic Data to Train Models for Identifying Hazardous Obstacles on Railway Tracks Learn more Sort by: Recently updated Nokia GenAl — GenAl Assistant to improve driver performance and Enhance Driver Safety to Reduce Road Accidents with an Analytics Platform Solution Learn more Orange GenAI — GenAI Assistant for field services, to improve quality and reduce costs of technical interventions by identifying faulty cables DEMO Learn more.

Scene 21 (11m 12s)

A yellow swirly pattern Description automatically generated.

Scene 22 (12m 36s)

Powering augmented engineering with Generative engine.

Scene 23 (13m 14s)

Data & AI Community in Capgemini Engineering. Community SharePoint.

Scene 24 (14m 7s)

Gen AI Frist level banner 1600x550px (1) TRAIN YOURSELF GenAI campus.

Scene 25 (14m 29s)

OVERVIEW OF GENAI EFFICIENCY GAINS FOR SW ENGINEERs, Engagement Managers… & more to come!.

Scene 26 (15m 0s)

Project lifecycle stage Effort % of total Use case Feature state: how it could look like Telecom Application Automation Type Suggested co-pilots COTS GenAI Tool Capgemini GenAI asset Custom GenAI asset Year1 in % Year2 in % Year3 in % Weighted productivity impact in entire lifecycle Pricing/ user/ year Project Management 2.69 % Project planning automation The GenAI System should update the MPP or Jira project plan based on the required changes provided as input. It should consider all dependencies, critical path, and then rebase the plan. Additionally, it should generate a change and impact summary outlining major impacts on milestones, new critical path, cost impact, manpower impacts, etc. Project Planning/ Replanning is a common challenge that often requires reestablishing the baseline. Inform stakeholders of cascading changes in the programme and update them on the corresponding implications. Assistance from an AI system in impact analysis and advice can save a significant amount of time. GenAI/AI Planview AdaptiveWork Unqork (https://www.unqork.com/) Copilot for Microsoft 365 NA NA 5% 15% 20% 0.54% $ 360.00 Analysis & Design 5.68 % AI/GenAI NA NA TBD 2% 5% 5% NA Development 41.54% This is is the combination of the development efforts across Analytics, Charging, Digital Opertaions, Mediation and Security. GenAI based coding copilot shall be leveraged for providing generating code, test cases, reverse engineering code etc. GenAI/AI E.g. GitHub Copilot, AWS CodeWhisperer GitHub copilot Developer Studio NA 7.50% 15.00% 15.00% 6.23% $ 468.00 Testing 19.91% Regression pack (Test case) selection GenAI system should take scope of work , Impact document and HPQC/Jira as input and capable of suggesting Test cases for the coverage. From test repository it shall analyze each area and suggest alrady indexed test cases for coverage AI/GenAI spaCy (https://spacy.io/) or NLTK (https://www.nltk.org/) in Python. NA UTP/ATLAS TBD 3% 8% 10% 1.99% NA.

Scene 27 (16m 8s)

[Audio] Competition is fierce. Clients have high expectation. The pace of innovation is fast. There is a learning curve for each of us and for the organization. We will learn as we go..

Scene 28 (16m 21s)

[Audio] We can do upskilling we can test… But to really learn we have to try with the clients. Leadership want you to price to win Let's take some well-understood balanced risk Let's not be over cautious.

Scene 29 (16m 37s)

[image] A rainbow colored object with black background Description automatically generated.