Über Maanik
Senior AI Engineer | RAG, Agentic AI & Business Automation
- Autonomous AI Agents: I build reliable agentic systems using LangGraph and CrewAI. These agents can perform complex tasks autonomously, such as conducting web research, managing CRM data, or drafting personalized communications, acting effectively as digital support staff.
- Secure RAG Systems: I develop custom Retrieval Augmented Generation (RAG) pipelines that allow you to securely query your internal documents (PDFs, Excel, Notion). I prioritize data privacy and GDPR compliance, ensuring your proprietary data remains secure using private vector databases.
- Rapid MVPs & Prototypes: For startups and internal business tools, I can build fully functional full stack applications (Next.js + Supabase + AI) in a very short timeframe. I handle the entire development cycle, providing you with a working product faster than a traditional agency.
Englisch
Muttersprachlich oder zweisprachig
Deutsch
Grundkenntnisse
Projekt- und Berufserfahrung
- WAMOCON GmbH,AI EngineerDIGITALAGENTUREN & IT-CONSULTINGAugust 2024 - Heute (1 Jahr und 10 Monate)Frankfurt, Germany● Architected an autonomous marketing ecosystem using LangGraph and Claude 4.5 Sonnet served via FastAPI to replace legacy workflows with specialized agents. Implemented a semantic caching layer with Redis to minimize API latency and token costs. The system generates an automated weekly campaign calendar detailing daily topics and posting times via Telegram, reducing strategy planning time by ~70% while maintaining response times under 200ms.● Engineered dual agentic pipelines for content lifecycle management using Gemini 2.5 Pro to distinctly separate "Generation" and "Review" workflows. Integrated native Guardrails in Vertex AI to filter hallucinations and Jira with Google Sheets for state management. This accelerated content production by 3x while enforcing a strict Human in the Loop (HITL) protocol to ensure 100% educational accuracy.● Led the architecture of the FIAE Learning Platform using Next.js and Supabase, wrapping the application in Docker for production deployment on Strato. Consolidated fragmented training tools into a centralized system with Type Safety using Drizzle ORM, delivering a scalable architecture that supports the entire apprenticeship lifecycle with optimized query performance.● Developed "HAI," the platform's integrated AI assistant, using PGVector and Fine Tuned Embeddings to handle domain specific vocational queries. Implemented Hybrid Search using Sparse and Dense vectors and dynamic query routing to select optimal models. This automation with low latency resolved over 90% of routine support tickets and significantly reduced the mentorship burden on trainers.● Implemented end to end ML solutions ranging from a master's thesis on defect prediction to production deployment. Utilized Random Forest and XGBoost to establish reliable models for decision making. Designed a comprehensive MLOps lifecycle with Drift Detection to improve model monitoring, ensuring seamless updates across containerized environments.
- Cognizant,Software and Data AssociateDIGITALAGENTUREN & IT-CONSULTINGDezember 2020 - Februar 2024 (3 Jahre und 2 Monate)Pune, Maharashtra, India● Led the deployment of machine learning models for semiconductor wafer fault detection using Random Forest and XGBoost. Achieved 87% accuracy by tackling class imbalance and conducting rigorous validation testing to ensure reliable defect identification in production.● Conducted Feature Importance Analysis to identify key variables contributing to faults. This provided the engineering team with actionable insights to optimize the manufacturing process and significantly improve quality control benchmarks.● Collaborated in an Agile team to customize Windchill PLM solutions using Java. Built robust backend services and integrated data pipelines to automate workflows between manufacturing sensors and the PLM dashboard.
- Leaf Innovations Pvt. Ltd.,Data Analyst InternDIGITALAGENTUREN & IT-CONSULTINGJuni 2019 - August 2019 (2 Monate)Delhi, India● Engineered a personalized Music Recommendation System using Unsupervised Machine Learning to deliver content based on customer listening habits.● Leveraged Collaborative Filtering and Market Basket Analysis to identify latent associations between genres. Achieved an 82% accuracy rate by optimizing model parameters for user preference prediction.● Developed a Tableau dashboard to visualize test results and user clusters. This enabled thorough analysis and supported informed decision making processes for the product team.
Empfehlungen
Sei die erste Person, die Maanik empfiehlt
Teile Deine Erfahrung aus der Zusammenarbeit mit diesem Freelancer.
Diese Freelancer passen auch zu Ihren Kriterien
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Ausbildung und Abschlüsse
- M.Sc in Data ScienceUniversity of Europe Applied Sciences2025Master's in Data Science
- Post Graduation in Data ScienceInternational Institute of Information Technology Bangalore2023Post Graduation in Data Science