Priyanshu Raj / AI Applications · Full Stack · Data

Practical AI.
Thoughtful
engineering.

I build AI-enabled applications and connect them to the data, APIs and interfaces that make them useful. My work combines LLM integration, full-stack development and a foundation in analytics and quality assurance.

Explore selected work Let’s connect
Based in Surat, India · Open to AI application roles · Remote preferred
AIRAG · Prompt engineering · Evaluation
DEVNext.js · TypeScript · Node.js
DATASQL · PostgreSQL · Analytics
Selected work / 01

Real problems.
Work you can explore.

Current application work and public repositories, with clear scope, technical context and honest project status.

8 projects shown

01 / BUSINESS AIAI & DATA

Inquiry → context
→ quotation.

OpenAIZoho CRMZoho Books
Work in progress

AI workflow integration

I contribute to an application that connects sales inquiries with item search, quotation preparation and follow-up workflows.

Read project details
Problem
Bring inquiry handling, customer context and quotation tasks into a connected application.
My contribution
Application development, API integrations, prompt engineering, GPT fine-tuning and model-response evaluation.
Current scope
The application is in development. I focus on validating integration behaviour and AI outputs before relying on the workflow.
02 / EXPLORATORY R&DAI & DATA

Sales assistance.
Multilingual voice.

Next.jsTypeScriptPostgreSQLpgvector
Research & development

AI sales assistant research

Exploring an AI sales assistant with multilingual voice calling and retrieval backed by PostgreSQL and pgvector.

Read project details
Problem
Investigate how an assistant can use relevant business context during sales conversations.
My contribution
Full-stack development and research into retrieval, prompts and voice-assisted workflows.
Current scope
In R&D. I am learning and implementing pgvector; the workflow is still being explored and developed.
03 / OPEN SOURCEAI & DATA

One workspace.
Multiple AI tasks.

TypeScriptElectronAI workflows
Public repository

OmniWorkspace

An open-source AI workspace for coding, research, SQL and data-analysis workflows, bringing application development and AI-assisted tooling together.

Explore repository (opens in a new tab)
Read project details
My contribution
I maintain and contribute to the project, including work committed through my acadify-admin account.
Engineering focus
Connect interface, application services and model-assisted workflows into a usable desktop experience.
Explore
The repository includes the source, architecture documentation and test configuration. Use the README for current setup instructions.
04 / MODEL QUALITYAI & DATA

Evaluate answers.
Investigate failures.

PythonLLM evaluationReference data
Public repository

AI evaluation framework

An organization project I contribute to for evaluating LLM responses against reference datasets and supplied context.

Explore repository (opens in a new tab)
Read project details
Problem
Make model-response assessment more structured and repeatable.
My contribution
Contributions to an evaluation-oriented codebase and its documentation.
Engineering focus
Reference-based evaluation, contextual response checks and clear inspection of results. Explore the evaluation examples and documentation on GitHub.
05 / CONVERSATIONAL AISOFTWARE

A conversation.
A connected application.

ReactTypeScriptNode.jsOpenAI
Public repository

Full-stack chatbot

A chatbot application connecting a React interface and Node.js backend with OpenAI integration.

Explore repository (opens in a new tab)
Read project details
Engineering focus
Connect a user-facing chat experience to backend model integration.
What it demonstrates
Full-stack application structure, API integration and conversational interfaces.
Explore
Review the source and README on GitHub. The repository brings the interface and backend integration together in one application.
06 / BACKEND SYSTEMSSOFTWARE

Structured APIs.
Clear application flows.

Node.jsExpressMongoDBREST APIs
Public repository

Donation-platform backend

A Node.js and Express backend for campaign and donation workflows, with MongoDB and JWT authentication.

Explore repository (opens in a new tab)
Read project details
Engineering focus
Build backend endpoints around application workflows, data persistence and authentication.
What it demonstrates
API development and the backend foundations used when connecting AI features to a wider product.
Explore
Review the public repository for implementation and setup details.
07 / DOCUMENT TOOLINGSOFTWARE

Documents in.
Usable text out.

PythonPDFMarkdown
Public repository

PDF to Markdown tooling

Python tooling for converting PDFs into Markdown, a useful starting point for preparing document-based information.

Explore repository (opens in a new tab)
Read project details
Engineering focus
Transform source documents into text that can be reviewed and processed.
Connection to AI
Document preparation is one step in retrieval workflows; useful retrieval still requires chunking, indexing and evaluation.
Explore
See the repository for supported conversion behaviour and setup.
08 / CHAT INTEGRATIONAI & DATA

AI conversation.
Inside Telegram.

PythonTelegramOpenAI
Public repository

Telegram AI bot

A Python chatbot project integrating OpenAI-powered conversation with Telegram.

Explore repository (opens in a new tab)
Read project details
Engineering focus
Connect model responses to a familiar messaging interface.
What it demonstrates
API-driven bot development and conversational workflow integration.
Explore
Explore the bot implementation and setup documentation on GitHub.

Current workplace applications are described at a high level. Explore the linked repositories for code and setup instructions. The first two projects are currently in development and R&D.

Expertise / 02

AI is one part.
The application is the whole.

I work across model behaviour, application code and data quality, so the pieces connect into a useful experience.

01 / AI APPLICATIONS

Context, prompts
& evaluation

Integrate models into workflows, refine instructions and inspect responses for relevance, consistency and unsupported claims.

RAGPrompt engineeringOpenAI integrationGPT fine-tuningAI evaluation

Currently developing: vector retrieval with PostgreSQL and pgvector.

02 / FULL-STACK DEVELOPMENT

Interfaces, APIs
& integrations

Build user-facing flows, connect backend services and debug application behaviour using AI-assisted development and code review.

Next.jsReactTypeScriptNode.jsExpressREST APIsPython
03 / DATA & QUALITY

Reliable inputs.
Useful insights.

Write SQL, clean and validate data, prepare analytics reports and apply a QA mindset to application and model outputs.

SQLPostgreSQLMongoDBData cleaningAnalytics reportingAPI testingRegression testing
01 / DEFINEUnderstand the task

Clarify the user need, available data and expected behaviour.

02 / BUILDConnect the pieces

Develop the interface, integrations and model workflow.

03 / CHECKTest the behaviour

Inspect data, edge cases and model-response failures.

04 / REFINEMake it easier to use

Iterate on prompts, application flows and documentation.

Experience / 03

A foundation in data.
A focus on AI applications.

My official role is Data Analyst. Alongside analytics responsibilities, I contribute to assigned AI development work.

Data Analyst

Octfis Techno LLP

June 2026 — Present · Surat, India

  • Write SQL queries and prepare analytics reports to support business decisions.
  • Clean and validate data, investigate inconsistencies and organize information for analysis.
  • Contribute to assigned AI application development, API integration, prompt engineering, GPT fine-tuning and model evaluation.
  • Use AI-assisted development to implement features, review generated code, debug issues and validate application behaviour.

AI Evaluation

Acadify Solution · Magic AI work

November 2025 — February 2026

  • Worked on prompt and response evaluation, reviewing model behaviour against task requirements.
  • Identified response-quality issues and documented feedback to support prompt refinement.

Backend Development Apprenticeship

Acadify Solution

August 2024 — January 2026

  • Worked on Node.js backend services, REST APIs and database-backed application workflows.
  • Built experience with MongoDB, PostgreSQL, integration debugging and data validation.

Web Solutions Intern

MarqSet

April 2024 — May 2024

  • Worked with WordPress, Shopify and Bootstrap on website implementation and responsive interfaces.

Quality Assurance

AppTestify

August 2021 — August 2023

  • Worked on functional, regression, mobile and API testing.
  • Used tools including Postman, Katalon and Repeato to inspect behaviour and identify defects.

Alongside these roles, I collaborate with friends on Krutsha AI across full-stack development, data cleaning and brand building. During college, I also worked as a freelance Technical Executive with Ireava.

About / 04

Curious about the problem.
Careful with the solution.

I’m an AI application developer with a background in data analysis, backend development and quality assurance. I enjoy turning an unclear requirement into a working flow: understanding the data, connecting the right tools and making the result usable.

My approach is practical. I use AI-assisted coding to move faster, then review the implementation, debug integrations and test what the application actually does. I bring the same care to model responses: clear prompts, relevant context and evaluation matter as much as the interface.

I’m continuing to deepen my work in retrieval and vector databases, while contributing to public projects and exploring AI engineering opportunities. My supporting experience includes technical content, search visibility and website analytics.

Build with purposeStart with the problem the application needs to solve.
Show the evidenceReview the behaviour, record findings and make the next iteration better.
01

AI application development

Model integrations, prompt workflows and user-facing AI features.

02

Full-stack implementation

Responsive interfaces, backend services and API integrations.

03

Data analysis & reporting

SQL queries, data preparation and practical analytics reports.

04

Evaluation & quality

Model-response review, application testing and integration checks.

Connect / 05

Have a useful problem
to work on?

I’m interested in AI Engineer, Generative AI Developer and AI application development opportunities. Remote work is preferred; I’m based in Surat, India.

Explore my code, connect on LinkedIn or email me with the role or project you have in mind.

Email mepriyanshubth6742@gmail.com ↗Connect on LinkedInPriyanshu Raj ↗Explore GitHub@Priyanshu845438 ↗