Brandon T Bande
Tech Community Lead | Speaker | Strategic Manager
Gweru, Zimbabwe
Actions
Since 2020 Brandon has been actively engaging with and speaking to the developer community on tech and non-tech related issues. Having presented to mixed and diverse audiences , Brandon delivers electrifying talks that aim to inspire , inform and engage audiences. His talks has ranged from physical and online workshop sessions to panels and keynotes. Some of the events he has been part of include DevFest Harare (Zimbabwe) , DevFest Kitwe and DevFest Lusaka (Zambia) and , BuildwithAI Bonny Island (Nigeria ), Mozilla Festival Zambia , PyCon Zimbabwe and PyCon Africa (South Africa).
As a community lead and builder , Brandon has contributed to the revitalisation of Google Developer Groups Harare , spreading and mentoring student tech communities through Applied Machine Learning Days Africa , Zindi and he previously served as a Microsoft Learn Student Ambassador and a vice-president for Students for Exploration and Development of Space Midlands State University.
Links
Area of Expertise
Topics
Using Gemma3:270m to build privacy-preserving local AI tools for low spec machines
Gemma3:270m is Google's smallest LLM yet and it can run on as low as 500 MB of RAM (Google's words not mine).
This beginner session shows how to build local AI agents using Gemma3:270m and Python
We will use Python to build an agent that scraps job requirements from different websites (For the demo I will use a personal API to store the data and retrieve that gives data that has been scrapped) , combines scattered information about me or anyone who would want us to use their documents , customize and personalise cover letters and resumes for each and every job.
The event will wrap up with the design considerations around the demo app and will center around :
1. The need for privacy for personal and organisations and why Gemma3:270m helps address that
2. Prompting techniques to ensure that the output is as desired
3. Use of guard rails and the other nitty gritty details to churn out better output
No you can't just build : How to use AI to create impactful projects not shelved weekend projects
Everyday we all use all kinds of AI for all kinds of things. We even tell ourselves that with AI we now have the power to build , scale and make impact yet we make none.
In this talk , Brandon , an M. Com in Strategic Management and Corporate student , tech community leader and trainee Business Analyst will dive into what makes a great solution in the AI bubble. The talk will talk about things often overlooked and key takeaways will include :
1. Identification of customers and customer needs : I too have built 'cool' and 'ground-breaking' stuff but I could not break into the market. Why? I built stuff no one was willing to pay for because I was too technical.
2. The hidden powers that kill most solutions : We will talk on things like how ethics , regulation and environment swallow up you idea before you even launch it. Some of the ideas we have will possibly create generational debt if regulation step up.
3. Documentation : Documentation takes time , to save more time and AI tokens. As a bonus , Brandon will take you through the basic things that he does before even opening the IDE.
4. Tips and hacks from my time with the tech community , academia and industry that has helped me prompt better.
What if the model never saw the image : Privacy-first Sign Language System with MediaPipe
Machine-learning engineering often focuses on building better models and collecting more data. But sometimes the more important decision is how the problem is represented.
In this talk, I share lessons from a personal project where limited labelled data led me to rethink a conventional computer-vision pipeline. Instead of feeding raw images into a model, I used MediaPipe Hand Landmarker to transform visual input into structured 3D hand landmarks, then normalized the representation to preserve relevant information while removing unnecessary variation.
This architectural choice affects data requirements, model complexity, privacy, and inference latency.
The talk explores a broader engineering principle:
Before reaching for a bigger model or more data, ask whether you can change the representation of the problem.
Build WITH AI , Build with AI , BuildwithAI : Google Toolkit to utilising AI in different scenarios
No you can't just build. You don't just get a lightbulb moment and start clattering on your keyboard expecting AI to work magic until you have a product.
This talk will explore 3 key definitions of Build with AI.
1. Building stuff making use of AI to do the heavy lifting - Using tools like Firebase Studio , Gemini Code Assist and even Opal to create MVPs and prototypes.
2. Using AI in an intentional and structured manner to build stuff. Using unlikely tools like Gemini Deep Research you can scrap the internet for problem research , project documentation , product design and so forth. I would add on tools like Stich for UI/UX designs and how to even use other tools out of the Google Ecosystem to finish the job ( I love Gemini but Kiro and Claude have shown to be exceptional to me)
3. Incorporating AI in the solutions that you are building. Ranging from AI Studio to Gemma to Vertex AI. Google offers tools to ensure that one can easily integrate AI in their solutions on edge or cloud. I will also touch on different offerings like Imagen , Veo3 and real time AI integration.
The key takeaways is that yes anyone can build stuff , but how do you leverage Google technology to build meaningful stuff. And what tools fit which scenario? Also what to consider (Risks around different tools ).
Easy as A, B, See : Integrating Gemini Pro model to Python for Computer Vision
It used to take years of learning to build computer vision applications but AI makes it simple. In less than 30 minutes I will build a CCTV system that is manned by AI to show how AI makes it easy to build and implement ideas. Using Python and AI Studio I will show how easy it is to utilize Gemini and share a few tips on building using Gemini.
Intentional Design for Low-Resource, Privacy-First Systems: How a 3 MB Payload Became 60 KB
What if building for constraints isn't about making a weaker system, but designing a better one?
In this talk, I share lessons from a personal project where a 3 MB payload was reduced to just 60 KB by rethinking how the problem was represented and where computation should happen.
Rather than relying on heavier models, larger servers, or sending raw visual data to the cloud, the project explores MediaPipe, WebAssembly (WASM), and on-device processing to build a system that is smaller, faster, more private, and better suited to environments with limited resources.
The talk uses this experience to explore a broader approach to engineering: design intentionally around the constraints of your users, devices, data, and environment. The key decisions made were around what information really needed to be kept/discarded , how much data we needed to transmit and how we could reduce user waiting time.
The goal isn't simply to make ML smaller. It's to ask a better question: What is the minimum system we can design that still solves the problem well?
PyCon Namibia
Workshop om designing delivery methods for AI solutions that focuses on combining Gemini API and Twillio for WhatsApp access.
AI Literacy Gweru Sessions
Understanding AI Beyond the Hype
Session 1 of a 6-Part Learning Series with American Space Gweru
Artificial Intelligence is everywhere—but what does it actually mean for everyday people?
Understanding AI Beyond the Hype is the first session in a six-part series hosted in collaboration with American Space Gweru, designed to help participants truly understand AI and use it confidently in their daily lives. This opening session cuts through the buzzwords and headlines to focus on what AI really is, how it works at a basic level, and why it matters to you.
In this session, we’ll explore the foundational building blocks of AI that are most relevant to everyday users—no technical background required. Participants will gain a practical understanding of concepts such as data, algorithms, machine learning, and common AI tools, while learning how these elements show up in the apps, platforms, and services they already use.
Whether you’re curious, cautious, or excited about AI, this session will help you build a solid mental model that prepares you for the rest of the series—and for an AI-driven world.
Building a Successful Career in Data & AI
Building a Successful Career in Data & AI was a practical and inspiring session for students and young professionals who want to break into the world of data science and artificial intelligence. Learn from top Zimbabwean AI practitioners as they share real career journeys, essential skills, and proven strategies for starting strong in this fast-growing field. Whether you're a beginner or exploring your next step, this event will equip you with clarity, direction, and opportunities to grow.
DevFest Mbarara 2025 Sessionize Event
Google Developer Groups Harare DevFest 2025 Sessionize Event
PyCon Africa
Vibe coding is okay when you are building for yourself. How about when a customer reaches out to you to build something for them? This is not often talked about when people talk about using generative AI for development. In this talk I will share on how I use the Google provided tools for MOST of my Generative AI assisted product development. Using brand identity and my own requirements gathering , I am able to use Gemini 2.5 for documentation , requirements analysis , the Stich for editable UI designs , AI Studio for prototyping and Cursor and Google's Jules for product building ( I could use Firebase Studio but as of now I cannot handle certain aspects with it). This talk will showcase my process and share a few oops moments that helped me finetune my process. I will wrap up with the common things to avoid before deployment for solutions that I have built for mobile and web - which includes obvious pointers to AI usage which might devalue my work.
This talk will be ideal for enthusiasts to mid level developers , resonate with senior developers and vibe coders , whilst its tone will help create room for discussions during tea breaks. In the same talk , I will share on how AI has helped me overcome impostor syndrome too by enabling me to build using the knowledge that I have whilst teaching me the need to understanding the basics to reduce technical debt associated with vibe coding.
BuildwithAI Bonny Island
Presented a virual technical talk on 'Retrieval Augmentation Generation and contextualising responses for personalised use in Gemini' for GDG Bonny Island.
Beyond the Classroom: Unlocking Career Potential
Delivered a career guidance in tech talk to GDGoC University as a guest speaker
Tech Tomorrow: Innovations Shaping the Future
7:05 PM - 7:25 PM: Keynote Speech 1
Topic: The Future of Artificial Intelligence: How It Will Transform Every Industry
Speaker: Brandon Bande
Discover how AI is revolutionizing industries by streamlining processes, enhancing customer experiences, and transforming human-machine collaboration.
Devfest Lusaka 2024 Sessionize Event
11th International Conference on Communication and Information Science
Research Paper Presentor
Deep Learning Indabax Zimbabwe
Panel Speaker on AI Governance and Ethics
Women Tech Makers (WTM) Onitsha Virtual Bootcamp Sessionize Event
DevFest Zambia 2023 Sessionize Event
DevFest Zimbabwe 2023 Sessionize Event
DevFest Zimbabwe 2022 Sessionize Event
Brandon T Bande
Tech Community Lead | Speaker | Strategic Manager
Gweru, Zimbabwe
Links
Actions
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top