Frame
Decision, users, sources, risks and success criteria.
00THE SIGNAL
Ibrahim Krayem · Generative AI engineer for financial, documentary and business workflows.
Turn complex information into decisions teams can use.
I design and build LLM applications, RAG systems and document workflows—from ingestion to evaluation and the user interface.
Currently developing generative AI applications at TALAN. PhD in Computer Science.
01THE PROBLEM
Start with the decision—not the model.
You do not need one more AI demo. You need a system that understands the right sources, exposes its reasoning path and produces a useful deliverable.
My research background brings disciplined modelling and evaluation. My engineering practice turns that discipline into software teams can operate.
- 01Clarify
- 02Connect
- 03Verify
02SIGNATURE CASE
From financial documents to a traceable SWOT workflow.
01 / SVI
Bpifrance · SVI agent · via TALAN
Modernisation of a financial chatbot ecosystem for investment and holdings monitoring.
- Problem
- Bring financial data, monitoring notes and user documents together in a structured SWOT analysis while preserving the expected format and a fluid real-time experience.
- My role
- I develop and evolve the SVI agent across the Python services and Angular interface.
Intervention — multi-format ingestion, sourced SWOT, SSE streaming and PowerPoint export.
- Ingest PDF and DOCX files and maintain an extraction agent for PDF, image, DOCX and PPTX inputs.
- Generate a SWOT with summary, sources and recommendations from financial and documentary context.
- Stream tokens, execution steps, tool calls and results through SSE.
- Export the analysis to PowerPoint and document the transition between Java, Python and MCP architectures.
Delivered transformationPDF · DATA → SWOT → PPTX
4 documented input formats: PDF, image, DOCX and PPTX.
Delivered result
Explore all projectsAn end-to-end conversational workflow that enriches LLM analysis with financial and documentary context and delivers reusable SWOT presentations.
03THE SYSTEM
Build the full chain. Keep every step observable.
Build
Retrieval, agents, document processing, APIs and interfaces.
Observe
Tracing, sources, execution steps and operational signals.
Evaluate
Quality checks, limits, feedback and controlled iteration.
- Financial AI agentDocuments and financial data become a sourced SWOT analysis and a reusable presentation.
- Document intelligenceLLMs support the analysis of business documents within the PICXEL project.
- RFP assistanceAn internal tool helps managers understand opportunities and prepare responses.
- Performance researchAn analytical model evaluates network-on-chip performance up to 500× faster in the reported research comparison.
04TRAJECTORY
Research rigour, applied to production AI.
The same question runs through every stage: how can a complex system be made measurable, understandable and useful?
Generative AI developer · TALAN
LLM applications, financial agents and document-analysis workflows.
R&D engineer · IRISA
Fast analysis of network-on-chip performance from real application traces.
Doctoral researcher · IRISA
A Python analytical model based on queueing theory, up to 500× faster than conventional simulations in the reported comparison.
PhD in Computer Science · University of Rennes
Research foundations in modelling, measurement and critical evaluation.
500×
Maximum speedup of the doctoral analytical model over conventional simulations in the reported comparison.
Certifications
- Microsoft Azure AI-900
- Professional Scrum Master I
- GitHub Copilot Certification
Hobbies
- Running
- Walking
Aqarih
Designing and building a bilingual real-estate platform enhanced by an AI-powered property search assistant.
Explore the personal project ↗05NEXT SIGNAL
A generative AI role or mission to make concrete?
I bring a research mindset, end-to-end engineering and a focus on traceable outputs to generative AI products.
PriorityA generative AI engineering role. I am also open to a short, clearly scoped mission.
Paris · Fluent French · Professional English · Native Arabic