Open to Principal, Staff & leadership roles Nairobi, Kenya (Remote)

Adioz ships backends and AI.

Full-stack AI engineer / Backend Systems & Production AI. Leads backend and AI engineering on a streaming platform's Java 21 and Spring Boot services: payments that never double-charge, auditable revenue sharing and real-time analytics. Builds AI the same way, owned from design to support: tool-using agents that make real, auditable changes, and small, quantized models that meet tight latency budgets on CPU.

Years experience
8+
Architectures designed
10+
Dependency CVEs cleared
150+
Automated tests
4k
adioz — agent.loglive
  1. requestplain-language change
  2. agent→ model on CPU · intent, slots
  3. agent→ typed tool · pages.update
  4. guardrail✓ gates passed
  5. auditapplied · operator on record
  6. requestplain-language change
  7. agent→ model on CPU · intent, slots
  8. agent→ typed tool · pages.update
  9. guardrail✗ gate failed · fail-closed
  10. rollbackchange reverted

01 Core Engineering Focus

AI as efficient as it is intelligent.

I build backend systems and take AI into production, and own each system from design to support: payments that never double-charge, agents that act within guardrails, and models sized to the latency budget and hardware they run on.

CODETRAINSHIPEST 2018
01

Tool-Using Agents

Agents built as production software: typed tools, fail-closed guardrails, and a named operator behind every change they make.

  • Typed tools
  • Guardrails
  • Auditable changes
02

Resource-Aware AI

Trains and distils models on GPU, then serves small, quantized models on CPU that meet tight latency budgets without wasted compute.

  • Distillation
  • INT8 quantization
  • CPU-only inference
  • Burn
  • Candle
03

Backend Systems

Java and Spring Boot services that charge exactly once, split revenue with a full audit trail and stream events into analytics, owned from API design to production support.

  • Java 21
  • Spring Boot 4
  • WebFlux
  • Kafka
  • ClickHouse
Service record
Robust
99% uptime held
Fluent
Five languages, front to back
Equipped
Vision, text and ranking models
Seasoned
8 years in the field

02 Tech Stack

Architectural toolkit.

The languages, data systems, and ML runtimes behind production AI.

Java & Spring Boot
Expert
Rust Systems & Tokio Async
Expert
Production AI (Distillation / INT8)
Expert
Kafka Streaming & ClickHouse
Expert
Cloud & DevOps (AWS / Terraform / Docker)
Advanced

Agentic AI

  • Tool-using and ReAct agents/evaluation
  • Typed audited tools
  • LangGraph
  • LlamaIndex
  • Model Context Protocol (MCP)
  • Retrieval (RAG)
  • Local SLMs

ML Engineering

  • Burn
  • Candle
  • PyTorch
  • Hugging Face
  • CLIP
  • Embeddings
  • LoRA/QLoRA/PEFT
  • vLLM/TGI/Triton/Ollama
  • MLflow/W&B
  • Model training and evaluation

Resource-Aware AI

  • Multi-teacher distillation into small students
  • Quantization-aware training (QAT) and INT8
  • Truncatable embeddings (MRL)
  • CPU-only inference
  • Zero-copy memory-mapped weights
  • SIMD scoring
  • Cached teacher passes
  • Accuracy-driven model sizing
  • Latency and memory budgets
  • Load profiling

Backend

  • Java 25 and Spring Boot 4
  • Spring WebFlux and R2DBC
  • Rust (Tokio, Axum)
  • Go (gRPC)
  • Python
  • REST, GraphQL and WebSockets
  • Microservices
  • Testcontainers

Frontend

  • Angular 22
  • Next.js
  • React
  • Astro
  • TypeScript
  • RxJS
  • Leptos/WASM
  • HTML5/SCSS

Data

  • PostgreSQL
  • MongoDB
  • SQLite
  • ClickHouse
  • Kafka
  • Redis
  • Tantivy (full-text search)
  • Qdrant

DevOps & Security

  • AWS
  • Terraform
  • Docker
  • Kubernetes
  • GitHub Actions
  • CI quality gates
  • SonarQube
  • Prometheus/Grafana
  • JWT
  • OAuth2
  • RBAC
  • Offensive Rust (Exploit PoCs, Custom Security Tooling, Low-Level Memory Security)

Working knowledge

  • Cloud TPU
  • JAX
  • DDP/FSDP
  • DeepSpeed
  • SageMaker
  • Vertex AI
  • Azure ML
  • CUDA

03 Experience

Production trace & impact.

adioz — career.trace5 spans · 2018 → nowlive

Apr 2026 — Present

Chief Technology Officer

BBR (Big Bold Red Agency) · Nairobi, Kenya

  • Engineering leadership: Set the architecture and review designs across Baze (Safaricom platform) and BBRaiN; plan and assign technical work to engineers; own all technical communications.
  • Platform modernisation: Led the upgrade of Baze's Spring services to Java 21 and Spring Boot 4, resolving breaking changes across every service; now leading the move to Java 25.
  • Operational excellence: Own runbooks, deployment change requests and incident response, with root-cause analyses.
  • Quality gates: Maintain the performance, smoke and end-to-end test suites, SonarQube checks and CI/CD pipelines that every release passes.
  • Java 21
  • Spring Boot 4
  • Rust
  • SonarQube
  • Snyk
  • Trivy
  • Terraform
  • CI/CD
  1. — Present

    Chief Technology Officer

    BBR (Big Bold Red Agency) Nairobi, Kenya

    • Java 21
    • Spring Boot 4
    • Rust
    • SonarQube
    • Snyk
    • Trivy
    • Terraform
    • CI/CD
    • Engineering leadership: Set the architecture and review designs across Baze (Safaricom platform) and BBRaiN; plan and assign technical work to engineers; own all technical communications.
    • Platform modernisation: Led the upgrade of Baze's Spring services to Java 21 and Spring Boot 4, resolving breaking changes across every service; now leading the move to Java 25.
    • Operational excellence: Own runbooks, deployment change requests and incident response, with root-cause analyses.
    • Quality gates: Maintain the performance, smoke and end-to-end test suites, SonarQube checks and CI/CD pipelines that every release passes.
  2. —

    Lead Engineer (AI and Backend)

    BBR (Big Bold Red Agency) Nairobi, Kenya

    • Rust
    • Axum
    • Burn
    • PyTorch
    • Java
    • Spring WebFlux
    • R2DBC
    • Kafka
    • ClickHouse
    • Redis
    • Sole engineer on BBRaiN: Designed and built the Rust AI engine behind search, recommendations and personalisation on Baze (Safaricom platform), deployed inside each client's own cloud (VPC).
    • AI in the binary: Serve vision, embedding and ranking models from one Rust service on CPU, with no GPU runtime, within a tight latency budget.
    • Payments and revenue: Built Baze's revenue-sharing engine from scratch, with an auditable payout workflow, and contributed to its payment, catalog and auth services in Java and Spring Boot.
    • Analytics and playback: Built Baze's analytics service for event ingestion and campaign attribution, and the playback events that flow through Kafka into ClickHouse and on to recommendations.
  3. —

    Junior AI Engineer

    Zone01 (Berrijam AI projects)

    • Go
    • Java
    • Python
    • gRPC
    • Docker
    • AI integration: Integrated ML models and LLM APIs into Go, Java and Python microservices.
    • LLM features: Built prompt, retrieval (RAG) and structured-output features for client-facing tools.
    • Data preparation: Cleaned, transformed and validated client datasets for modelling and analytics.
    • Prototype to production: Turned validated AI prototypes into services with MLOps pipelines.
    • Evaluation and documentation: Tested model outputs against baselines and documented APIs.
  4. —

    Fullstack Engineer (Java and Go)

    Zone01 Kisumu, Kenya

    • Java
    • Go
    • gRPC
    • Kafka
    • Service performance: Cut inter-service overhead by 30% by moving services to Go and gRPC.
    • Streaming pipelines: Ran Kafka pipelines with sub-second ingestion for real-time monitoring.
    • Team standards and mentoring: Set code-quality and DevSecOps standards, reviewed code, and mentored apprentice engineering teams.
  5. —

    Backend Engineer (Go and Java)

    Upwork Remote, international clients

    • Go
    • Java
    • Spring Boot
    • WebSockets
    • Angular
    • Docker
    • Real-time systems: Built Go (REST, WebSockets) and Java Spring Boot backends for startups.
    • Full-stack delivery: Integrated Angular frontends with Go and Java backends.
    • Reliability: Deployed Dockerized services with auto-scaling and monitoring, holding 99% uptime, and resolved production incidents for clients.
GitHub activityread
Contributions
14,794 since
Current streak
2days11 – 12 Oct 2026
Longest streak
29days4 Sep – 2 Oct 2026

Contributions per year

  1. 20181
  2. 2019Gitea · Bitbucket0
  3. 2020Gitea · Bitbucket0
  4. 2021Gitea · Bitbucket0
  5. 2022540
  6. 2023440
  7. 20242,678
  8. 20255,828
  9. 20265,307

Years at 0 went to client work in private, confidential Gitea and Bitbucket repositories, which GitHub does not count.

Languages across owned repositories

  • Rust32.9%
  • Python13.3%
  • Jupyter Notebook10.0%
  • Java9.7%
  • TypeScript8.0%
  • C6.6%
  • Other19.4%

04 Featured Projects

Systems built for scale.

Baze (Safaricom platform)

Safaricom's video and music streaming platform on web, Android, iOS and TV. Built its revenue-sharing and analytics services, designed the payments rail for M-Pesa and carrier billing, led the upgrade to Java 21 and Spring Boot 4, and delivered features on Baze Web.

  • Java 21
  • Spring Boot 4
  • WebFlux
  • R2DBC
  • Kafka
  • ClickHouse
  • Redis
  • Next.js

Payments that never double-charge, and revenue shared with a full audit trail

VybStreamz

The entertainment hub powered by Safaricom. Built content moderation with BBRaiN AI for its operations team, integrated Safaricom single sign-on into the CMS, and configured BBRaiN as a separate tenant with its own catalogue and rules.

  • Next.js
  • React
  • TypeScript
  • SSO
  • BBRaiN

BBRaiN moderates its content and powers its search and recommendations

BBRaiN

One AI engine for search, recommendations and personalization across all media: a Rust inference engine, a GPU training factory and a signed model hub that delivers models into each client's own cloud (VPC), run through an MLOps loop from training to drift monitoring. Replaced Recombee and Algolia on Baze (Safaricom platform).

  • Rust
  • Axum
  • Burn
  • CLIP
  • Tantivy

About 200,000 lines of Rust, backed by about 4,000 tests

Hisiya

AI sentiment intelligence for African markets. Reads brand sentiment in Swahili, Sheng, Pidgin and Amharic languages.

  • Swahili
  • Sheng
  • Pidgin
  • Amharic

Turns conversations into a brand-health score, trends and benchmarks

05 Workflows

Systems, one step at a time.

How the systems I build move requests, money, data and models: each one an engraved diagram you step through, every hop captioned with what happens there and what can go wrong.

14 parts · 11 steps

The resource-aware MLOps loop

Large teacher models are distilled into one small student, trained for INT8 and served on CPU, shipped as signed bundles into each tenant's own cloud, and watched in production until drift starts the loop again.

  • PyTorch
  • Burn
  • Rust
  • MLflow
  • Ed25519
  • Prometheus

Step through it

24 parts · 16 steps

The event-driven order path

An order on a Kubernetes platform is checked at the edge, recorded once with its event, carried by Kafka through services that reserve the stock and authorise the card, with the payment captured only when the order ships, and traced from checkout to dispatch.

  • Kubernetes
  • Helm
  • Istio
  • Envoy
  • Postgres
  • Debezium
  • Kafka
  • RabbitMQ
  • Redis
  • ClickHouse
  • OpenTelemetry
  • Prometheus

Step through it

32 parts · 15 steps

A subscription streaming platform

Sign-in, the home screen, a checkout and a play request from the web, phone and TV apps, each through its own backend for frontend, with Kafka carrying the events, change data, heartbeats, activity and logs behind them.

  • Kubernetes
  • Helm
  • Istio
  • Envoy
  • OIDC
  • Postgres
  • Debezium
  • Kafka
  • Flink
  • RabbitMQ
  • Redis
  • ClickHouse
  • OpenTelemetry
  • Prometheus
  • Grafana

Step through it

From peers

12 parts · 10 steps

The MLOps loop

A model is built, checked, released in stages and watched in production. When the live data drifts away from what it was trained on, an alert starts the loop again.

  • PyTorch
  • Optuna
  • MLflow
  • Prometheus

Contributed by Ruchita Singla

Step through it

06 Contact

Let’s take it to production.

Open to Principal, Staff and engineering leadership roles, and to teams that need backends and AI running in production, on their own hardware and in their own cloud.