Need better buffer with clients in terms of delivery planning - let clients know that we work 7 hours a day or need contingency. Need to account for 0.5 day technical debt, internal calls, admin etc. Make use of Marina to communicate this to client.
PMs need to ensure that sufficient team members are allocated to meet project timelines
Importance of following design
Allocating resources to a team who are inefficient derailing
Enforcing that clients abide by scope of work
Ensure that contracts have terms to limit scope of work to a time budget
Figure out how to ignore the urge to add 'extra' features
Recognising with your task is going to take longer than planned size
How to communicate scope changes about task size
Developers should recognise out of scope requests and offer the client the opportunity to create a new contract / contract extension
Delivering quality work, not just doing the bare minimum
Use your PM to proxy communicate with clients when needed
Develop skills in sizing tasks. Always break down tasks into atomic units. You get a psychological advantage of closing off lots of small tasks.
Learn to hack your own brain to keep yourself motivated and be able show client regular progress
We should match the skills of our team to projects requiring their skills and not assign tasks where they will have to spend too much time learning new technologies not familiar to them.
Quoting for digitising work: Break the country / region into 30km2 and calculate around 2 days per block. Then add another 2 days for QA per regions.
Project kick offs:
Drawio meeting where we have a high level but technical project overview discussion and expectations setting
Where project scope is unclear, there should be a budgeted activity for scope clarification and the scope should be firmed up with the client before starting actual work. Marina should be in the loop on this so that we can hold the client accountable to limit the scope to what we have agreed.
Clients should be given parameters for engagement: fixed number of iterations e.g. for review, deadlines for design requirements, clear communication as to when activities have been budgeted for.
Plan for learning curves in new technology
Build in expert time in projects to swiftly bypass blockers
Escalate problems efficiently so that you dont stick on things for days that a team mate could solve in minutes
Level up our SDLC maturity level to allow us to provide more robust solutions
Migrate to new infrastrcture Q1 MVP, Q2 Migrations
Optimised local environment managed by devops so devs can be productive within a day
Operational Areas:
Kartoza inhouse systems (should be behind VPN e.g. NC)
Kartoza common infrastructre (e.g. postgis.kartoza.com) that can be client facing
Developer tools / developer local env
Training deploy and scrub platforms
Bespoke Client Hosted Applications
Generic Client Hosted Applications
Kartoza Hardware Inventory
Communication plan for clients with the upcoming migrations
Communication plan for developers on processes and workflows for SDLC
Application catalogue for everything we manage with details of who the client is, what the billing regieme is, who the internal owner is, where the documentation is etc. Workflow for new infrastructure needs to include maintaining the application catalogue. Every application and resource deployed should be tagged with an app id which is tracable back to the app catalogue.
Can we use ERP next for the above?
Goal to have no orphaned infrastructure
Standardising architectures as much as possible
Need to be able to specify per-region deployments where customer constraints require it
Observability needs to be built into the stack from the ground up (dashboards, alerting, log views)
Devops environment needs to be documented from the ground up
Implementation of planner sheet integrated with ERP Next in timesheet app
Burn down charts in timesheet app
Strong focus on team leader roles
Focus on building our team communication skills (not only in the technical sense but in being proactive communicators)
Internal training e.g. Seabilwe presenting basic GIS skills to non GIS staff
Succession planning: To ensure that the company does not have reliance on a single person
Put in place mechanisms to onramp developers efficiently into our devops stack e.g. developer VMs
Proposal development:
Internal testing: plan to include testing as a standard and integral part of every project, budgeted as a line item. Testing blowback (developer work that results in lots of failed tests) should reflect in developer KPI's
Functional testing in CI
Regression testing in CI
Workflow testing in CI (using Playwright)
Integration/Deployment testing using DevOps / ArgoCD
Manual testing with a test script
Two weeks at end of contract for sign off
Client management: we need to ensure that our contracting and project plans manage expectations of clients and that contracts halt until each phase is signed off
Proposal development should factor for 7 hour days not 8 in the project plan
Look for standardisation in project technology:
Devops SDLC
Django
Postgres / GIS
ReactJS
MapLibre
Vector Tiles
Establish spaces for the company to hang out together
We need a plan for security:
Securing staff computers - policies and communication with staff about best practices -
local disk encryption
have I been pwned checks
require password managers
strong password requirements
etc.
Security at network level:
Internal systems (NC, Gdrive, ) requireing YubiKey
Move internal systems into VPN
Share incentive scheme in place in 2023
Preferential shares eligible for dividends and bonuses
Shares vest when the company lists on the stock exchange (not applicable for us?)
Establishment of Kartoza.eu and managing impact on Kartoza.za
Team meet up
Sales handovers - standardize the handover of new work coming in
