Abhishek Saikia
Systems and Software Developer
India
About
Junior Developer with a passion for building systems and applications. Currently seeking opportunities to further develop my skills in a professional setting.
Education
Jorhat Engineering College
Computer Science
Experience
-
- SDE Intern @ Rablo.inLucknow, Uttar Pradesh - RemoteSummary:
- Developing and maintaining web applications using Next.js, Scss, and HTML. Collaborating with the team to implement new features and fix bugs.
Responsibilities:
- Built and optimized production-grade UIs using Next.js, Redux, improving load time by about 20% through code-splitting, memoization and SSR tuning.
- Integrated automated mailing workflows using Nodemailer and Node.js, reducing manual overhead in onboarding flows and improving user activation roles.
- Designed REST API workflows and Redux state management, cutting client-side response latency by about 15% and ensuring responsive UIs across all viewports using SCSS and Tailwind CSS.
Achievements:
- Resolved 50+ bugs and Git merge conflicts, improving stability across CI/CD build pipelines using structured branching workflows.
- Collaborated in Agile/Scrum sprints with design and backend teams to ship 6+ production features in TypeScript to 1K+ active users.
-
- AI Intern @ NIT SilcharSilchar, Assam - RemoteSummary:
- Using BioBERT, a pretrained biomedical language representation model for biomedical text mining, to understand the symptoms of patients and predict the disease.
Responsibilities:
- Built a disease prediction model using NLP (TF-IDF, tokenization) and classification algorithms (Random Forest,SVM) in Python with scikit-learn, achieving 98% validation accuracy.
- Developed a doctor recommendation engine using cosine similarity on patient symptom vectors, improving match precision over keyword-based filtering.
- Built Python backend services for model inference, data preprocessing, and real-time prediction APIs using Flask.
Open-Source
-
PR - #2093 Open Source Contribution @ Go AthensSummary:
- Identified an observability gap: Athens exposed HTTP-level OpenCensus metrics but lacked proxy-layer cache and upstream fetch metrics, limiting operator visibility into cache effectiveness and upstream reliability.
Contributions:
- Implemented three additive OpenCensus metrics in Go under pkg/observ: cache lookup total (Counter; labels: hit/miss + cache type), upstream fetch total (Counter; labels: success/failure), and upstream fetch duration seconds (Histogram, exponential buckets).
- Designed with low-cardinality labels and zero API/storage changes; integrates with existing Prometheus, Datadog, and Stackdriver exporters. Added unit tests for all three metric paths
-
PR - #3288 Open Source Contribution @ go-swaggerSummary:
- Identified a runtime panic in the schema parser where comment-based enum declarations assumed a non-empty type slice, causing crashes during spec generation when the type was uninitialized.
Contributions:
- Fixed a slice out-of-range panic by guarding access to sv.current.Type[0] and safely handling cases where the schema type was not yet initialized.
- Preserved existing parsing behavior by passing an empty type to enum parsing logic when unavailable, ensuring backward compatibility.
- Added a regression test covering the full Run() path with a swagger:model struct using comment-based enum declarations, reproducing the crash before the fix and validating correctness after.