Project: Recipe NER Pipeline.
Objective: Extract structured entities (AMOUNT, UNIT, FOOD, DESC) from 1 million raw recipes using Python and spaCy v2.
The Reality: The client's provided JSON training data is broken. The Doccano annotations are botched, riddled with overlapping entity spans and amateur labeling mistakes.
Execution Plan:Scrub the garbage data. Fix the overlapping span errors. Force the spaCy model to train correctly. Deploy a pipeline that actually works.

oftware helps recruiters process candidates faster by parsing resumes, comparing them against a job description, scoring the fit, and drafting outreach automatically.The workflow automates resume screening by extracting structured candidate information, evaluating alignment with the role, and generating recruiter-ready follow-up content.

Or even simpler: Resume in->data extracted->matched to job->candidate scored->outreach draft generated.

created llm apps and notebooks Reformatting the Excel file. Cleaning and organizing the data into the required format. Completing the analysis and final output for review.

created an NER that can label the following, Person, Names, Phone numbers, emails, address, IDs, links, money amounts, company, Name, dates.

created a program that can pull name, email, and phone number extracted from pdf and Word files within the same folder

Ai app that does following: 1) Extract/ Scrape text from URLs 2) Preprocessing of data 3) Fetch results from Openai 4) Save results into csv