About Asotele
Asotele is an open economic forecasting platform for financial institutions and businesses operating in emerging markets. The first market is Nigeria — chosen because the problem is biggest there, the data is hard, and if it works there, it works anywhere.
Why this exists
Banks and asset managers operating in emerging markets face a problem that doesn't exist in OECD economies: the analytical infrastructure that supports good investment and risk decisions is structurally weaker. Bloomberg, Refinitiv, and S&P Capital IQ are priced for global institutional desks; they cover emerging markets thinly and at price points that exclude domestic mid-tier banks, regional asset managers, and businesses making consequential FX and treasury decisions.
The data exists — it's just unfused. Central bank releases, exchange data, food and fuel prices, news sentiment, employment, cross-border flows. Asotele's job is to fuse them in real time and run economic models on the result, then surface the output as forecasts a bank's treasury, a fund manager's portfolio committee, or a corporate CFO can act on.
What Asotele is, technically
Three layers:
- Data layer — 17 daily scrapers, currently focused on Nigeria: oil, official and parallel FX, CBN monetary policy, NGX equities across 9 sectors (34,349 historical points), food and fuel prices, cement, employment, telecom subscribers, news sentiment via three NLP models, Google Trends consumer signals, World Bank macro indicators, cross-border USDT flows.
- Forecast layer — three econometric engines: ARIMA, Markov regime-switching with GARCH(1,1) volatility classification (boom/stress/crisis), VAR + Granger causality multi-variable analysis.
- Intelligence layer — fine-tuned open-weight foundation LLMs (Qwen 3.6 35B-A3B Mixture-of-Experts, Gemma 4 26B-A4B) producing natural-language briefings, scenario analyses, and Q&A grounded in the structured forecast outputs.
Read the full architecture page for the technical detail.
Built from Nigeria, designed to travel
Nigeria is the first market because the founder is Nigerian, the economic conditions reward analytical rigor, and the data infrastructure is genuinely hard (three concurrent FX regimes, sparse and intermittent official statistics, dense informal-economy signals). The same architecture replicates to Ghana, Kenya, Egypt, South Africa, Pakistan, Vietnam, Argentina, or any emerging market with comparable data availability — in weeks of pipeline reconfiguration, not years of platform rebuild.
Why open-source
The core platform is open-source under Apache 2.0 because credibility is the scarce resource. A bank's chief data officer can't deploy a black-box forecasting model into a regulated risk workflow without auditable methodology. Open-source code, openly published evaluation benchmarks, and transparent model weights are the cost of entry for that conversation. The business model is therefore commercial integration, deployment, and support — not gated software.
Who's behind this
Apex Grid Technologies is a Nigerian-registered company based in Lagos. The team also operates Finovamax, a banking software product — Asotele integrates natively into Finovamax deployments and is offered standalone to institutions that don't use Finovamax. Founder: Francis Oyakhire.
Two audiences, one engine
Asotele's outputs are served via two distinct surfaces — same forecast engine underneath, different presentation layers:
- For financial institutions (paid tier, live). REST API, Finovamax-embedded, or private VPC deployment. Banks, asset managers, and treasury desks integrate Asotele forecasts directly into existing risk, trading, and research workflows. This is the commercial layer that funds the platform.
- For businesses and founders (free tier, in development). A web and WhatsApp chat interface in English, Hausa, Yoruba, Igbo, and Nigerian Pidgin. The graduate starting a drone factory, the market trader managing dollar exposure, the SME owner planning a capex round — all use the same forecast engine via a free conversational surface, subsidized by institutional revenue and grant funding.
This is intentional structural design: the addressable Nigerian SME market cannot bear Bloomberg-tier pricing, but it's exactly the audience where good economic forecasting changes outcomes most. Banks pay for the infrastructure; small businesses get the outputs. Apex Grid wins on both fronts and Nigeria gets economic intelligence its citizens were previously priced out of.
Current status
- Data pipeline operational since November 2025; daily briefings produced since April 2026
- ARIMA + regime-switching forecast engines in production
- VAR + Granger causality engine activating end-May 2026
- Foundation-model fine-tunes (Qwen 3.6, Gemma 4) pending compute capacity
- First pilot bank conversations beginning Q3 2026
- SME-facing free tier (web + WhatsApp, multilingual) launching Q4 2026 pending LINGUA Africa decision