HARSH TRIPATHI · AI ENGINEER & FOUNDER · PARIS, FRANCE / REMOTE · AVAILABLE OCTOBER 2026

I build and ship AI systems that automate real-world workflows.

AI engineer working across LLM applications, agentic workflows, RAG, backend systems and machine learning — from architecture to production deployment.

Capabilities: AI Engineering · Agentic Systems · LLM Applications · Backend · ML
System Architecture Flow: INPUT STREAM → FASTAPI GATEWAY → REASONING & RAG AGENTS → DATA & SERVICES → PRODUCTION OUTCOME

02 / What I Build

  • AI Agents: Systems that reason, use tools and execute multi-step workflows. (Tool Calling, Autonomous Loops, State Machines)
  • LLM Applications: Production applications using structured generation, retrieval and tool use. (RAG, Embeddings, Structured Output, Prompt Orchestration)
  • AI Automation: AI systems connected to APIs, browsers, databases and business workflows. (Playwright, Webhooks, CRM Sync, Async Queues)
  • Backend Systems: FastAPI services, APIs, databases, asynchronous workflows and containerized deployments. (Python, FastAPI, PostgreSQL, Redis, Docker)
  • Machine Learning: Predictive models, feature engineering, data pipelines and applied ML systems. (Python, Scikit-learn, SQL, Pandas, Feature Pipelines)

03 / Featured Case Studies

01 · OCTAGRAM — AI-Powered Business Automation Platform

Role: Founder & Technical Lead | Badge: PRODUCTION · 6+ CLIENTS | Website: https://octagramai.com

Summary: Founded and technically led a technology company building digital platforms, autonomous AI automation engines and business software for commercial clients.

Problem & Approach: Service businesses suffer from disjointed manual processes for lead discovery, research, qualification, and CRM entry. Designed bespoke automation engines combining browser agents (Playwright), LLM reasoning pipelines, CRM webhooks, and modern web applications.

Tech Stack: Python, FastAPI, Playwright, PostgreSQL, Redis, React, Docker.

02 · FOREO — AI-Powered Advertising Operations

Role: AI Contract Engineer | Badge: PRODUCTION · CLIENT

Summary: An AI-powered platform designed to automate the lifecycle of digital advertising operations — from campaign planning and product data processing to creative workflows, publishing, monitoring and optimisation.

Problem & Approach: Global product lines required extensive manual hours coordinating briefs, creative variations, catalog synchronization, and live channel deployment. Engineered a modular AI pipeline with dedicated task-specific agents orchestrated via FastAPI backend services, vector search (RAG), and containerized microservices pushing directly to ad platform APIs.

Tech Stack: Python, FastAPI, Docker, Vector Search, Multimodal AI, PostgreSQL, Redis.

03 · FOREO — Multimodal AI / RAG System

Role: AI Intern · Team Lead | Badge: PRODUCTION · TEAM LEAD

Summary: A multimodal AI application combining image classification, dense retrieval, and LLM-generated responses grounded in structured knowledge bases.

Approach: Led a team of four engineers building a dual-stream ingestion pipeline extracting joint visual and textual embeddings into FAISS, integrating Ollama local inference and cloud LLM APIs via FastAPI.

Tech Stack: Python, FastAPI, PyTorch, FAISS, CLIP, Ollama, Docker.

04 · FOREO / Analytics Research — Influencer ROI Prediction

Role: ML & Backend Developer | Badge: PRODUCTION · ML SYSTEM

Summary: An ML system using multi-platform business data to estimate creator campaign performance and support marketing decision-making.

Approach: Built SQL ETL pipelines aggregating past campaign records, engineered engagement velocity and authenticity features, and trained Scikit-learn regression/ensemble models served via FastAPI.

Tech Stack: Python, Pandas, NumPy, Scikit-learn, SQL, FastAPI, PostgreSQL.

04 / Experience & Education

OCTAGRAM — Founder & Technical Lead (2026 — Present, Remote) [PRODUCTION · 6+ CLIENTS]

Founded and lead a technology company building digital platforms, AI automation systems, and business software for 6+ clients.

FOREO — AI Contract Engineer (2026 — Present, Paris / Remote) [PRODUCTION · CLIENT]

Building AI-powered advertising operations, agentic workflows, backend services, integrations, and automation systems.

FOREO — AI Intern · Team Lead (2025 — 2026, Paris, France) [PRODUCTION]

Led a team of four engineers building multimodal AI applications, RAG systems, backend services, and ML workflows.

TRAINITY — Data Analyst Intern (2023 — 2024, Bangalore, India)

Business analytics, automated reporting, data processing, and SQL/Python workflows.

ATIC-FRANCE — Co-founder / AI Developer (2022, Paris, France)

NLP pipelines for speech-to-text extraction and automated video summarisation.

Education & Credentials

BITS Pilani: B.Sc. Computer Science (2022 — 2026) · GPA: 9.27 / 10

IBM: Data Science Professional Certificate

05 / Engineering Process

  1. 01 — Understand: Turn the business problem into measurable technical requirements.
  2. 02 — Design: Define architecture, data flows, APIs and failure modes before implementation.
  3. 03 — Build: Develop modular systems with clear interfaces and validation.
  4. 04 — Deploy: Containerize, integrate, monitor and validate production behaviour.
  5. 05 — Iterate: Use real usage and system feedback to improve the product.

06 / Technologies & Infrastructure

AI / ML: Python, PyTorch, Scikit-learn, LLMs, RAG, Embeddings, Computer Vision

Backend: FastAPI, REST APIs, PostgreSQL, Redis, Docker, Pydantic, SQLAlchemy

AI Infrastructure: FAISS, Ollama, Vector Databases, Model Serving, Uvicorn

Development & Ops: Git, GitHub, Linux, Docker, CI/CD, PyTest

07 / Contact & Connect

Email: rdht.work@gmail.com

LinkedIn: https://linkedin.com/in/rdh-tripathi-325752307

GitHub: https://github.com/rishabhprojects-stack

Location: Paris, France (Open to remote & onsite roles)

Availability: Full-time roles starting October 2026