Mar 2023 – Mar 2025 · 2 yrs

• Worked on agentic use-cases of KnowledgeAI to help customers better
• Conducted extensive experimentation, service development and realtime serving
• Created KnowledgeAI (https://www.observe.ai/real-time/knowledge-ai), Observe's flagship RAG offering from scratch
• Worked on Realtime ML capabilities involving Generative AI (GenAI).
• Show-and-tell paper accepted at Interspeech 2023

Jun 2021 – Mar 2023 · 1 yr 9 mos

• Part of the Natural Language Understanding Team working on building and deploying models that understand transcripts of free flowing conversations.
• Involved in not just research but also service development, containerisation and deployment on cloud with considerations for scaling via loadtests and more.

Feb 2021 – May 2021 · 3 mos · Bengaluru, Karnataka, India

• Adding to Intelligent Tutoring Systems (ITS) capabilities by exploring Deep Learning (Transformer based) approaches to capture a student's knowledge state to improve the learning experience for millions of students.
• Recommend the best content to students based on the above information.
• Tech stack: PyTorch, Attention-based models, LSTM, RNN, ConvGNNs etc.

Aug 2017 – Jul 2019 · 1 yr 11 mos · Bengaluru Area, India

• Designed and developed novel algorithm to structure any semi-structured document.
• Published workshop paper in leading document analysis conference.
• Tech stack: Statistical machine learning, Gibbs sampling, natural language processing, PDF parsing

Jun 2017 – Aug 2017 · 2 mos

• * Worked on the DBPedia knowledge graph

Jul 2016 – Dec 2016 · 5 mos · Bengaluru Area, India

• Worked on natural language processing and machine learning problems being faced at belong.
• Worked on concepts such as named entity disambiguation, topic modelling, text classification, sentiment analysis, high recall classifiers etc.
• Developing end-to-end machine learning pipelines; right from building datasets to training and deploying machine learning models.
• Tech stack: Named Entity Recognition and Disambiguation, Sentiment Analysis, Topic Modeling, tensorflow

May 2016 – Aug 2016 · 3 mos

• Gensim is one of the most popular open source machine learning and natural language processing library in python in the world.
• I added the topic coherence pipeline to Gensim to evaluate all kinds of topic models on the human interpretability of the topics they produce, automatically. This process used to be otherwise a tedious manual process.
• Tech stack: python, Latent Dirichlet Allocation, Topic Coherence, Data Visualisation

Aug 2015 – May 2016 · 9 mos · BITS-Pilani, Goa

• BITS F.C. is a professional football club based out of Birla Institute of Technology and Science, K.K. Birla Goa Campus and plays in the GFA Goa League 1st Division. BITS F.C. is the only university team in India that plays in a professional league. BITS F.C. also participates in different university level tournaments all over India.

Aug 2015 – Mar 2016 · 7 mos · BITS-Pilani, Goa

• The core team is responsible for the management of the club ranging from people-management to finances and events.

Jun 2015 – Dec 2015 · 6 mos

• Searock India is the largest semi-professional band competition in India. I was responsible for managing the New Delhi city elimination round held at Hard Rock Cafe, Gurgaon.

May 2015 – Jul 2015 · 2 mos

• Developed a unified portal for room booking and food billing at Vedanta. Also developed the visit Vedanta website. Used C#, HTML, Javascript and frontend frameworks such as materialize.