Syed Alam Shah Bukhari

AI Engineer specializing in production-grade RAG, LLM extraction, and NLP systems.

Based in: Hyderabad, PakistanMost recently: AI Engineer Intern, Feature Lead (COA Pipeline), PM Accelerator, Attuned AIExperience: 1 year

Hire me

Experienced AI Engineer with a proven track record of shipping hybrid retrieval systems and structured LLM APIs for international startups. He has successfully led feature teams to deliver high-accuracy evaluation harnesses and real-time speech analysis pipelines.

Experience

AI Engineer Intern, Feature Lead (COA Pipeline)

PM Accelerator, Attuned AIRemote, USAinternshipJun 2026 – Aug 2026

Summary

Promoted to feature lead mid-internship; owned retrieval architecture end-to-end and coordinated two engineers' parallel workstreams to deliver the COA AI pipeline by Demo Day. Built a hybrid RAG retrieval system combining keyword search with OpenAI text-embedding-3-small embeddings and pgvector cosine similarity over a 30-scenario knowledge base, with automatic fallback logic. Designed and shipped an LLM extraction API producing structured JSON across 29 behavior codes and 65+ trigger categories using Next.js/TypeScript and Groq/OpenAI. Built a 100-case evaluation harness achieving ≥92% behavior-code accuracy, 100% safety-escalation recall, and <5% false-confidence rate; resolved production blockers across PostgreSQL RLS/RPC and Docker/Supabase.

What I did

Built a 100-case evaluation harness to formalize the evaluation process.

Results

Achieved ≥92% behavior-code accuracy in a 100-case evaluation harness.
Achieved 100% safety-escalation recall.
Maintained <5% false-confidence rate.
Delivered the COA AI pipeline by Demo Day.

RAGOpenAI text-embedding-3-smallpgvectorPostgreSQLNext.jsTypeScriptGroqOpenAISupabaseDockerRLSRPC

AI/ML Developer Intern

OlloRemote, Toronto, CanadainternshipSep 2025 – 2026

Summary

Built a real-time speech-analysis pipeline using Hugging Face Whisper, FastAPI, and JavaScript for filler-word detection and live inference visualization, owning model integration, API design, and production-level prototyping.

What I did

Owned model integration, API design, and production-level prototyping.
Implemented transcript-based filler-word detection for disfluencies like "um" and "uh" using post-processing.

Hugging Face WhisperFastAPIJavaScript

Skills

Technical

RAG

GitHub

Python

OpenAI Embeddings

LLMs

Git

LangChain

LLM Evaluation

PostgreSQL

Supabase

Docker

FastAPI

TypeScript

JavaScript

pgvector

ChromaDB

Hugging Face Transformers

Whisper

Next.js API Routes

Inngest

scikit-learn

Overlapping chunks

Linear Algebra

Probability & Statistics

Machine Learning

Independent project creation

Automated debugging workflows

Projects

RAG Pipeline, PDF Question Answering System

Developerpersonal

Built and deployed an end-to-end document Q&A system with PDF chunking, Gemini embeddings, ChromaDB retrieval, and context-grounded generation; resolved cloud deployment issues involving ChromaDB compatibility, libffi, and API timeouts.

PythonLangChainChromaDBGemini APIStreamlitGemini embeddingsReusable test workflows

Visit the site View the code

Freelance Admin Agent

personal

Built an end-to-end system to track invoices and payments from unstructured email text, defining requirements and architecture independently. Used a Strands agent with Claude via Amazon Bedrock for extraction and deterministic Python/SQLite for financial logic.

Strands agentClaudeAmazon BedrockPydanticPythonSQLite

Education

Bachelor of Science

Mehran University of Engineering & Technology (MUET)Artificial IntelligenceAug 2024 – Dec 2028

Recognition

Certifications

Certified AI Engineer

PM Accelerator2026

Foundation: Introduction to LangChain (Python)

LangChain Academy2026

Social impact & activities

Pie & AI Ambassador

DeepLearning.AIorganizerJun 2026 – Present

Founded and led Hyderabad's first Pie & AI chapter event, 'Introduction to RAG Pipelines & LLMs,' live-demoing a self-built RAG system to 50+ registered attendees.

Introduction to RAG Pipelines & LLMs

Pie & AI HyderabadworkshopJun 2026

Live-demoing a self-built RAG system to 50+ registered attendees.

Hire SyedReach out about a role, a contract or a conversation.For recruiters

Syed's twin is AI, it can make mistakes.

More people on Kanvis

Browse more people Create your own Kanvis