Zohaib Tesneem

Responsible for leading a team of developers to build modern APIs. Working throughout most stages of the software development lifecycle from design, build, smoke test and deployment.

Responsible for solutioning a key document management process to support Virgin Media O2's objective to work more efficiently throughout their network roll-out lifecycle stages. This involved migrating and standardising the data model into the cloud from a tightly-coupled user interface. The solutions were presented to all stakeholders (internal and external).

Architect for an AI-powered Cognitive Assessment Tool for a client. The project scope was to implement a cloud-hosted, serverless, intelligent prioritisation tool. This received inbound requests for support via a webform / voice recording to a call centre, and prioritised the requests based on sentiment analysis and keywords. Role:

· Designing the application architecture for the serverless, AWS-based solution

· Primary upward interaction with direct supervisor – i.e., Product Owner, Delivery Lead, and InfoSec Lead

· Network architecture design experience

· Working closely with the InfoSec lead to find the right balance between security and performance

· Technology Stack: ReactJS for the front-end, AWS-based backend consisting of API Gateway, Lambda functions, S3 buckets, DynamoDB for the database and IBM Watson for the AI service

· Some key security decisions made:

o AWS KMS to encrypt S3 buckets, DynamoDB tables, Environment variables and to provide application-level encryption

o Defining the strategy for key rotation policies

o Role-Based-Access-Controls defined and documented

o IP whitelisting to only retrieve requests from a webform

o IP whitelisting to UK IP addresses from CloudFront for the UI dashboard to reduce surface of attack

· Network design decisions:

o Defining public and private subnets with route table rules

o The use of a NAT Gateway in a public subnet to prevent inbound traffic

o The use of AWS VPC Endpoints to connect Lambda functions to S3 buckets and DynamoDB and other services

o Application-level encryption to protect Personal Identifiable Information (PII) data being sent from Lambda functions

o Defining IAM policies to follow the principle of least privilege

o The use of TLS v1.2 to protect data in transit

o Defining alerts and proof of concept created to send errors and updates to policies via email

o Users can only sign-in to the UI dashboard through Multi-Factor-Authentication

o Admins can only create accounts for users in AWS Cognito

o Password Policies defined and documented for both UI dashboard access and the AWS console

o Key security decisions were made by following Policy-56 – Accenture Cloud Security Configuration Standard

· A data flow diagram to highlight the flow of data from a webform, the UI dashboard and the FTPS server (where the audio files were extracted from) to the AWS backend

· Produced an automated design for data deletion within 7 days to meet GDPR requirements

· Operational dashboard design (including alerts) – this included a full analysis on what could be used currently vs. what additional services may be needed (e.g., CloudTrail) to meet missing requirements

· Documentation – High-Level Design and Low-Level Design – these needed to be completed in a short period of time to ensure the DevOps team could progress with their tasks

· Cost calculations for any additional services that may be needed – such as SMS for Multi-Factor-Authentication

· Estimation of development efforts provided to the delivery lead on new design decisions that needed to be implemented

· Risks and concerns raised with approaches/suggestions where good design practices, performance or security were sacrificed

The Athlete Genome was a machine learning solution that bridged the gap between performance and sentiment to support British triathlon athletes be better prepared for the Tokyo 2020 Olympics. Following go-live, a requirement came in to improve the performance of the application. Role included:

· Working closely with a performance tester to understand higher latency cause

· Implemented a caching solution to reduce cost and latency when making calls to the application tier

· Suggested database indexes be used to reduce the time taken for data retrieval

· Produced a report which was reviewed by the product owner

· Once implemented by the development team, end-to-end testing was performed by the performance tester to look for improvements; the result was that the latency issue had been resolved