AI Engineer · Python systems

Open to AI Engineer & ML Engineer roles

Models answer.I engineer what happens next.

01 Role: AI Engineer
02 Evidence: 05 inspectable projects
03 External review: 03 merged upstream PRs
04 Core: Python · FastAPI · AI integration

Flagship evidence Volyx Lens View architecture and source

I build Python systems for retrieval and context engineering, multimodal input, model routing, backend reliability, and evaluation.

View selected work View CV

Interactive system field The model is one node. The engineering is the field.

context plane control plane

01 / context

Select relevant text, screen, voice, and meeting context before a model sees it.

Python backends

Retrieval + context engineering

Multimodal + voice AI

Evaluation before claims

Selected evidence / 05

Selected engineering work

Five inspectable builds showing context retrieval, model integration, async APIs, realtime state, and honest evaluation.

Direct AI evidence

Multimodal AIContext engineering

Volyx Lens

Status: Active pre-release
Ownership: Independent project
Evidence: Source · architecture · CI

A privacy-oriented macOS assistant combining selected screen context, microphone input, meeting audio, local OCR, relevance-ranked retrieval, transcription, and multiple AI providers. I hardened its context boundaries, provider routing, consent controls, tests, and release checks. Public builds are ad-hoc signed test builds, not notarized releases.

Implemented controls: context exposure is constrained by selected inputs, consent controls, and explicit provider boundaries.

Screen, speech, and meeting context
Ranked retrieval with explicit boundaries
Provider routing and consent controls

Supporting engineering

Real-time systemsBackend engineering

Noughtline

Status: Public demo
Ownership: Independent project
Evidence: Live demo · tests · CI

A React and Express multiplayer game with signed guest sessions, server-validated moves, reconnect handling, an append-only transaction ledger, and one-time crediting for verified Paystack references.

Verified behavior: moves are validated by the server and verified payment references can be credited only once.

Python integration

Async API integrationPython

VirusTotal Telegram Bot

Status: Source available
Ownership: Independent project
Evidence: Source · implementation

An async Telegram bot for VirusTotal v3. It hashes locally before upload, follows the large-file flow, backs off on rate limits, and caches reports for interactive details.

Flow Hash locally → query → cache report

Python integration

API aggregation prototypePython

Solana & Ethereum Wallet Analyzer

Status: Public prototype
Ownership: Independent project
Evidence: Live demo · tests · CI

An async Telegram bot that validates Solana and Ethereum addresses, combines four RPC and market-data sources, caches requests for five minutes, and paginates Solana holdings in chat.

Flow Validate → gather → merge → cache

Archived experiment

ML evaluationPython

Football Predictor

Status: Archived evaluation
Ownership: Independent project
Evidence: Evaluation report · tests · CI

An archived XGBoost and Poisson forecasting experiment. Across 1,140 rolling-origin test matches, it reached 53.77% accuracy versus a 56.70% bookmaker benchmark, so I did not present it as a betting product.

Conclusion Evidence changed the product decision

External review / 03 merged

Merged upstream work

Focused fixes accepted into established open-source projects after maintainer review.

Swipe or use the controls to view all three

Contribution 1 of 3

FastStreamMerged into main · Jul 2026

FastAPI compatibility

Added support for FastAPI 0.140 compatibility after changes to its Dependant interface.

View merged PR #2961

OpenAI Agents SDKMerged into main · Jul 2026

Realtime retry handling

Fixed retry handling for pre-response WebSocket server errors in realtime sessions.

View merged PR #3991

CalkitMerged into main · Jul 2026

Scoped project checks

Ensured single-item runs skip checks for unrelated subprojects.

View merged PR #1028

Engineering method

From model call to
reliable system.

At VolyxAI, an independent product effort, I apply the same four-part method across language, vision, and speech systems.

01
Integrate
Connect language, vision, and speech providers behind explicit interfaces.
provider layer
02
Retrieve
Assemble ranked, bounded, source-aware context for the task.
context layer
03
Validate
Use schemas, deterministic checks, retries, and human approval.
control layer
04
Evaluate
Measure quality, latency, cost, and failure behavior before claims.
evidence layer

See what I'm building at VolyxAI

Engineer profile

What I build

I build Python systems around language, vision, and speech models: provider integrations, retrieval and context assembly, FastAPI services, async workflows, structured outputs, and evaluation.

AI engineering: retrieval · context engineering · multimodal · voice · structured outputs
Models & providers: Azure AI Foundry · OpenAI · Anthropic · Gemini · DeepSeek · Deepgram
Languages: Python · SQL
Backend & evaluation: FastAPI · asyncio · REST · n8n · scikit-learn · XGBoost

Open channel

Hiring an AI Engineer for
reliable model-backed systems?

Open to AI Engineer and ML Engineer opportunities.

Email me about a role CV LinkedIn GitHub