Loading profile…
Loading profile…
Saeed Bidi — Machine Learning Engineer · Point · London, England, United Kingdom.
Direct contact
1 verified contact method on file — work email. Masked until you unlock.
Free account, no card required. Your first unlock is on us.
Profile updated Nov 2026 · Sourced from public professional profiles · Legitimate-interest B2B basis
Or explore more contacts in this database — from 45 credits.
About
Forward Deployed / Applied AI Engineer (PhD). I find where AI creates real value inside established businesses, then build and ship it. For the past couple of years I've taken GenAI products from whiteboard to production in regulated finance — scoping the problem with stakeholders, building the backend, and making sure it actually gets used. Recent work at Point: an agentic briefing platform that turns portfolio, CRM and market data into compliant, meeting-ready client briefs in under a minute (live across client workflows); the business's first MCP-based product; and a GDPR-compliant document-intelligence platform for scanned financial PDFs. I work the way forward-deployed teams do: embedded with clients, owning delivery end-to-end, comfortable with ambiguity, focused on outcomes over activity. Stack: Python · FastAPI · LangGraph · MCP · RAG · GPT & Claude · Azure
Experience
Forward Deployed / Applied AI Engineer
Point · London Area, United Kingdom
Jan 2025 – Present
Owning the architecture, build and production reliability of enterprise LLM platforms used in live financial-services workflows. Embedded with clients end-to-end, from discovery to deployment. Core domain: multi-currency portfolios, FX, securities and investment books of record (IBOR). ● Point AI — Advisor Intelligence & Briefing Platform: Led the end-to-end build of an agentic system that unifies portfolio data, CRM signals and live market news into compliant, meeting-ready client briefs in under a minute — replacing hours of manual prep. Live as multi-tenant SaaS on Azure Container Apps. Multi-model LLM orchestration in LangGraph, Azure Service Bus batch pipeline, dual-JWT auth, full CI/CD. ● Secure LLM Integration (MCP): Designed and shipped the business's first MCP-based product, now live with clients — a production MCP server (FastMCP) letting external LLM clients (Claude Desktop, ChatGPT) securely query internal financial APIs in natural language, with authentication, scoped access and validation. ● FusionDocs — GenAI Document Intelligence: Built a GDPR-compliant platform extracting structured data from complex scanned financial PDFs (bank statements, asset-manager reports) via OCR + hybrid deterministic/LLM pipelines with schema validation, monitoring and auditability; PII de-identification before every LLM call on EU-hosted Azure OpenAI. ● Hybrid RAG — Financial Knowledge Retrieval: Built a hybrid retrieval system (vector + BM25) over large internal financial databases with grounding enforcement, embedding versioning and incremental re-indexing. Stack: Python · FastAPI · Flask · AsyncIO · Pydantic · LangGraph · LangChain · MCP/FastMCP · Docker · Azure (Container Apps, Functions, OpenAI, Service Bus, Key Vault) · CI/CD · vector DBs (Chroma, FAISS, pgvector)
Senior Data Scientist
Queen Mary University of London · London, England, United Kingdom
Jan 2023 – Jan 2025
● Designed and deployed reproducible ML pipelines processing health data for 60,000+ individuals, owning the full lifecycle from ingestion and modelling to operationalisation. ● Introduced automated testing, containerisation and CI/CD, reducing deployment failures and improving reproducibility. ● Applied time-series and survival analysis to longitudinal clinical datasets; delivered outputs directly to senior stakeholders and policy partners. ● Contributed to research published in Nature.
Education
Skills
Related professionals · Point