Adilkhan Alikhanov

@adilkhan

AI Engineer | LLM agents, RAG, Evals, Python

Backend and AI engineer based in Astana. I build LL-powered products and the backend systems behind the.

About

At ForteBank I built ForteExpert, an internal RAG ssistant for the first-line support team: hybrid retieval with FAISS and BM25 merged by RRF, a custom reanker tuned to banking data, and a FastAPI backend dployed inside the bank's closed network. I took it fom prototype to official release. Now I work on the ackend of an NGO ERP platform and its donor platform around 900 API endpoints, e-signature integrations uilt for reliability, role-based access control, andan LLM agent with tool calling over the platform's on API. Stack: Python, Django, FastAPI, Celery, PostgeSQL, Redis, Docker, FAISS, OpenAI and Anthropic API.

Career

Experience: 2 years

Experience

Backend & AI Engineer

SEN Digital (eNGOhub)

Dec 2025 — Present

ERP platform for NGOs plus a donor platform. LLM agent with tool calling over the platform's own API (custom agent loop on OpenAI/Anthropic, conversation memory, RBAC-scoped actions, voice via Yandex SpeechKit). E-signature subsystem with durable inbox, transactional outbox, lease-based locking and a recovery worker. RBAC with 9 roles and 100+ permissions. 900+ API endpoints, 130+ models, 7000+ tests. Python, Django, DRF, Celery, Redis, PostgreSQL, Docker.

Data Scientist

ForteBank

Nov 2025

ForteExpert: internal RAG assistant for the first-line support team over 4700+ chunks across 10+ internal bank systems. Hybrid retrieval (OpenAI embeddings + FAISS, BM25, merged with RRF), custom 9-signal reranker with banking context, sibling-chunk expansion, token-budgeted context. FastAPI backend with Redis chat history, Docker/Gunicorn inside the bank's closed network via DMZ proxy. Took it from prototype to production release.

Education

MSc Data Science

Nazarbayev University

2024 — 2026

MSc Data Science (2026). Also BSc Computer Science with a minor in Mathematics at Nazarbayev University (2024).