✦ Updates

Binu
Shefield
Shifani

AI/ML Engineer · Full-Stack Developer · Master's Researcher

5+ years building production AI systems at Cognizant — from LLM document pipelines to enterprise automation. Now building an on-premise agentic AI decision-support system for LP-DED aerospace repair as an M.Sc. thesis at GKN Aerospace / Högskolan Väst.

5+
Yrs Industry
8+
AI Projects
M.Sc.
AI & Automation
GKN
Thesis
B
Binu Shefield Shifani
📍Trollhättan, Sweden
🎓M.Sc. AI & Automation — Högskolan Väst
✈️Thesis Researcher @ GKN Aerospace
💼Former AI/ML Associate @ Cognizant · 5+ yrs
🌐English (Fluent) · Swedish (Learning)
Areas · Evidence
Agentic AI & LLMs
380-entry ChromaDB KB · dual-query RAG · 6-agent LangGraph pipeline @ GKN
Computer Vision
IoU 0.61 marine debris · YOLOv11 garment QC
ML Pipelines
OCR + GenAI doc extraction · Azure ML workflows
Full-Stack & .NET
5 yrs C# · Angular · Docker · CI/CD at Cognizant
LangGraph PyTorch RAG · VectorDB Azure ML LP-DED YOLOv11 TensorFlow Docker · CI/CD
03

Projects

Fetching latest from GitHub…
01

Technical Stack

🤖
AI / ML
PyTorchTensorFlowScikit learnHuggingFaceLangChainLangGraphOpenCVYOLOv11CLIPTransformersDeepLabV3
🧠
LLMs & Agents
RAGAgentic AIChromaDBVectorDBNLPOCRMultimodalPrompt Eng.GroundingDINO
☁️
Cloud & DevOps
Azure MLAzure CognitiveAzure FunctionsDockerCI/CDGitSpeech SDK
💻
Languages
PythonC# / .NETTypeScriptAngularSQLASP.NET MVCREST APIs
🔒
Security
SonarQubeSASTDASTBlack DuckSecure SDLC
⚙️
Robotics & Industrial
ABB RAPID codeArUco markerTwinCAT PLCLP-DEDMatlab
02

Experience

GKN Aerospace
Jan 2026 — June 2026 📍 Trollhättan, Sweden completed
Role
Master's Thesis Researcher — AI Decision Support Systems Jan 2026 – Present
What the System Does
Built a fully on-premise agentic AI decision-support system for LP-DED pre-deposition repair planning of flight-critical aerospace components — turbine blades, fan blades, compressor cases in Ti-6Al-4V and Inconel 718
Constructed a 380-entry ChromaDB knowledge base from peer-reviewed LP-DED literature, structured across 3 layers: single-bead parameters, multi-layer deposition, and repair case studies
Designed a novel dual-query Layer 1 retrieval strategy — two parallel queries with result merging and deduplication — addressing a known RAG limitation where single queries miss entries due to varied parameter terminology across publications
Implemented a 3-condition LLM grounding check before any recommendation is surfaced — ensuring suggestions are traceable to retrieved literature, not hallucinated or extrapolated beyond the knowledge base
Orchestrated a LangGraph multi-agent pipeline: Retrieval Agent → Consistency Verification → Suggestion Agent+ Critic Agent → Intent Classifier(feedback Agent) — each stage independently validating before passing results downstream
Implemented round-based knowledge base expansion: engineer-approved parameter sets are written back into ChromaDB through a two-phase approval process, making the system progressively more capable through validated operational experience
Deployed fully air-gapped — no cloud infrastructure, no proprietary databases — addressing aerospace traceability and IP requirements that rule out standard cloud LLM solutions
Addresses a direct industry gap: existing LP-DED repair approaches are entirely operator-dependent, with knowledge lost on personnel changes and no mechanism for systematic capture — this system provides a traceable, knowledge-grounded alternative
LangGraphChromaDBFAISSRAGAgentic AILP-DEDTi-6Al-4VInconel 718On-Premise LLMPythonAerospace
Cognizant Technology Solutions
Jan 2019 — Aug 2024 · 5+ Years 📍 Chennai, India Former
Progression
Associate — AI/ML & Full-StackDec 2022 – Aug 2024
Programmer Analyst — AI/ML & .NETDec 2020 – Nov 2022
Programmer Analyst TraineeJan 2019 – Dec 2020
Key Contributions
Built multilingual LLM pipeline with Azure Translator — enabled product rollout across 6+ language markets without engineering rework per locale
Led OCR + generative AI extraction pipeline for unstructured documents — replaced a manual review step that consumed ~20 hrs/week across the team
Integrated Azure Speech SDK for voice-driven automation — eliminated a category of ~15 repetitive operator interactions per shift in client workflows
Automated document classification and routing with UiPath + ML models — cut average document handling time from ~8 min to under 1 min per item
Architected C# .NET Windows Services for enterprise automation — removed dependency on 3 scheduled manual processes that previously required overnight operator oversight
Built ASP.NET MVC document management platform — improved document retrieval UX and reduced support tickets from end users within first month of deployment
Containerised .NET services with Docker on Linux — cut deployment time from hours to under 15 minutes and standardised environment parity across dev/staging/prod
Introduced SAST/DAST/SonarQube into SDLC — identified and resolved 40+ high-severity vulnerabilities before production across multiple release cycles
PythonC# .NETAzure MLLLMsOCRAngularTypeScriptDockerCI/CDUiPathSpeech SDKSonarQube
04

Education

Completed
M.Sc. AI & Automation
Högskolan Väst, Trollhättan, Sweden
Sep 2024 – Jun 2026
Machine LearningDeep LearningRoboticsBig DataImage ProcessingAutomation SystemsCybersecurity
Completed
B.E. Electronics & Communication
PSNA College of Engineering & Technology, India
Aug 2015 – Apr 2019
ElectronicsSignal ProcessingCommunication SystemsEmbedded Systems