LinkedIn | GitHub | Bellevue, WA | awagh2410@gmail.com
ABOUT ME
Hello! I’m Akanksha—an information scientist by diploma, a GenAI explorer by ardour, and a curious human at all times chasing “what’s subsequent.” By day, I’m constructing chatbots, decoding buyer sentiment, and optimizing campaigns. However once I’m not wrangling code or plotting forecasts, you’ll most likely discover me modifying movies, exploring new sports activities, writing weblog posts about life, or looking down new locations to eat. This weblog is the place tech meets persona—just a little area the place I share what I’m constructing, studying, experiencing, and infrequently obsessing over.
I’m actively in search of full-time roles in Knowledge Science and Machine Studying Roles. Should you’re hiring or know of a chance, that is my email: awagh2410@gmail.com and LinkedIn. Let’s join! Thanks 🙂
EDUCATION
Stevens Institute of Know-how, NJ
- Graduated with a Grasp’s diploma in Knowledge Science in Might 2025.
- Programs: Optimization for Knowledge Science, Utilized Machine Studying, Time Sequence Evaluation, Deep Studying, Statistical Strategies, Augmented Intelligence and Generative AI, Advertising Analytics, Dynamic Programming and Reinforcement Studying, and Generative AI for Enterprise Purposes.
Savitribai Phule Pune College, India
- Graduated with a Bachelor’s diploma in Data Know-how in June 2023.
- Awarded as ‘The Finest Outgoing Scholar 2023 (Excellence in curricular, co-curricular, and extra-curricular).’
WORK EXPERIENCE
- Knowledge Analyst (Advertising) at Compass Group, USA
February 2024-Might 2025
- Constructed Actual-Time Dashboards
I created interactive dashboards utilizing Energy BI to visualise marketing campaign information in actual time. These dashboards helped advertising and marketing and management groups monitor how varied campaigns had been performing throughout completely different social media platforms. I included visible components for sentiment developments, person engagement, and publish attain to offer a fast, actionable overview. - Analyzed Social Media Sentiment and Engagement
Utilizing sentiment evaluation, I evaluated buyer notion of the model on companies supplied. I recognized which varieties of posts resonated most with completely different audiences and supported in adjusting content material tone, timing, and platform methods accordingly. - Utilized Pure Language Processing (NLP) to Suggestions Knowledge
I labored with unstructured textual content information from buyer evaluations, survey responses, and suggestions types. By making use of NLP strategies, I extracted insights about person satisfaction, recurring points, and general sentiment developments, which fed into our buyer retention methods. - Ran A/B Testing for Marketing campaign Optimization
I helped design and consider A/B checks to check variations in digital advertising and marketing campaigns — from topic traces to content material layouts and call-to-action placements. Primarily based on the outcomes, I really helpful which variations carried out higher and why, offering data-backed steerage to the advertising and marketing supervisor. - Queried and Cleaned Advertising Knowledge
I wrote SQL queries to retrieve information from CRM programs and automation platforms. I cleaned and pre-processed this information to make sure it was analysis-ready, typically working with a number of information sources that wanted to be merged or filtered for relevance. - Tracked Key Advertising KPIs
I tracked and reported on important metrics like impressions, click-through charges, bounce charges, conversion charges, and extra. I additionally helped outline what “success” regarded like for particular campaigns by aligning with stakeholders on key efficiency indicators (KPIs). - Collaborated with Cross-Purposeful Groups
My work typically sat between technical and non-technical groups. I collaborated with entrepreneurs, designers, and enterprise managers to translate enterprise objectives into measurable outcomes, after which offered insights in a digestible, visible format.
2. Machine Studying Intern at Stevens Institute of Know-how
Might 2024-August 2024
- Designed and Deployed a GPT-4 Chatbot for IT Help
I led the event of an AI chatbot constructed on GPT-4 and LangChain, particularly designed to help with college IT helpdesk queries. The purpose was to offer college students and employees with an automatic, conversational interface that would reply widespread technical points — from forgotten passwords to software program troubleshooting. - Built-in LangChain for Context Consciousness
One of many key challenges was sustaining context in multi-turn conversations. I used LangChain’s capabilities to provide the chatbot reminiscence — permitting it to comply with up intelligently, reference earlier interactions, and preserve the dialog pure and coherent. - Automated Context Switching for Higher Question Dealing with
The chatbot was enhanced to deal with a number of varieties of queries throughout the similar session. I applied logic to detect when a person modified subjects or launched a brand new subject mid-conversation, and educated the bot to reset or pivot its responses accordingly with out breaking movement. - Constructed Actual-Time Monitoring Dashboards
I developed dashboards that tracked chatbot efficiency throughout varied dimensions — together with how lengthy it took to reply, how typically customers got here again, how correct the solutions had been, and whether or not the bot wanted to escalate the problem to a human agent. - Carried out Sentiment and Development Evaluation
By analyzing chat logs and person suggestions, I carried out sentiment evaluation to grasp person satisfaction ranges. This helped establish patterns in what varieties of questions the bot dealt with effectively vs. the place it wanted enchancment. - Collaborated Throughout Tech and Help Groups
This venture required working carefully with each builders and IT assist employees. I translated real-world helpdesk ache factors into conversational flows and iterated on design primarily based on suggestions from each customers and assist brokers. - Improved Person Expertise and Diminished Guide Load
The chatbot finally acted as the primary level of contact for a lot of IT-related queries, which lowered the load on the employees and improved response instances. It served as a 24/7 assistant that supplied constant, correct assist with out requiring a stay individual always.
3. Knowledge Engineer Intern at Tata Motors Restricted
January 2023-June 2023
- Constructed Finish-to-Finish Knowledge Pipelines for Engagement Metrics
I developed an information pipeline utilizing Python with Pandas and PySpark to assemble and course of person engagement information. The purpose was to scrub, rework, and consolidate data from varied inside instruments right into a centralized PostgreSQL database, making it simpler for groups to entry and analyze constant information. - Developed Backend APIs Utilizing Django REST Framework
To assist information entry for downstream instruments like Energy BI, I constructed RESTful APIs utilizing Django and Django REST Framework. These APIs uncovered processed information securely and effectively, permitting enterprise customers to attach their dashboards instantly with out counting on guide exports or scripts. - Designed Dashboards for Behavioral Insights
I created interactive dashboards in Energy BI to visualise person habits information — similar to how workers interacted with inside platforms or companies. These dashboards enabled completely different departments to trace developments, establish bottlenecks, and enhance function adoption. - Collaborated with IT Service Desk for Help Knowledge Evaluation
I labored with the inner IT assist staff to discover helpdesk information, significantly specializing in service requests raised by workers. I used Python, SQL, and Energy BI to carry out exploratory evaluation, perceive recurring ache factors, and visualize request developments throughout time and departments. - Carried out Multi-Instrument Knowledge Exploration
My work concerned integrating a number of instruments — utilizing SQL for querying, Python for evaluation and transformation, and Energy BI for storytelling and visualization. This full-stack method helped me see how uncooked information turns into actionable insights at an enterprise stage. - Centered on Optimization and Knowledge Accessibility
Each activity, from pipeline improvement to API publicity and dashboard creation, was pushed by the purpose of bettering accessibility, velocity, and readability for decision-makers. I aimed to scale back friction in information workflows so non-technical groups may make knowledgeable decisions while not having engineering assist each time.
AWARDS & PUBLICATIONS
- Published a paper titled “Self-Generated Soccer Highlights” in IEEE
Developed a soccer match summarization pipeline utilizing an Inflated 3D CNN for motion recognition and movement detection. Utilized the SoccerNet dataset and OpenCV to establish key occasions — objectives, substitutions, and penalties — and generate correct, concise spotlight clips. - Recipient of the “Best Outgoing Student” Award — 2023
Honored as the highest graduating pupil for excellent efficiency throughout lecturers, co-curricular initiatives, and extracurricular contributions. The award acknowledged a well-rounded mix of educational excellence, management, and energetic engagement in campus life.
CONTACT
I’m actively in search of full-time roles in Knowledge Science and Machine Studying Roles. Should you’re hiring or know of a chance, let’s join!
