Shebagi Mitra@shebagi
5
Events attended
2
Submissions made
Bangladesh
1 year of experience
About me
Technical mentor for lablab.ai. Interested and building projects in AI, Machine learning, Data science, engineering and coding in general. Feel free to reach out for any help regarding our hackathons!
I built with
Cohere🤝 Top Collaborators
.png&w=640&q=75)
.png&w=640&q=75)
.png&w=640&q=75)
🤓 Latest Submissions
.png&w=828&q=75)
Aviation Safety Evidence Engine
Aviation Safety Evidence Engine Aviation investigators synthesize long, conflicting PDFs where untraceable AI summaries fail. Our dashboard lets analysts click a component on an aircraft schematic to build a bounded evidence pack and generate a verified briefing: documented findings, unknowns, and next checks. Unanalyzed parts stay gray so safety is never implied. Tech Stack & Split Evolus prepares data via REST (Document to data: 52.8 credits/32.3s; Summary: 11.72 credits/5.2s), creating an unmerged overview labeled "generated, not verified." Gemma 4 31B IT (AMD Developer Cloud) provides open-weight reasoning suitable for private deployment, while deterministic code verifies every claim. The 7-Part Harness Code governs what reaches the screen: (1) Bounded packs with leakage guards and contradiction searches; (2) Strict prompts requiring passage IDs and verbatim quotes; (3) Injected JSON schemas with one repair retry; (4) A deterministic substring verifier that flags—never hides—failed quotes as "needs review"; (5) Provenance pulled strictly from stored metadata; (6) No model calls on empty packs; and (7) Full audit logs for a replay-only demo. Results & Limits Tested on an Alaska Airlines 737 gear collapse (NTSB DCA23FA417, cause withheld) across 0–44 passages (up to 15.7k tokens). Showing the schema raised first-try parsing from 0/3 to 3/3, and the verifier passed 13/16 citations. Runs identified the fracture mechanism but missed the root cause (overhaul grinding heat damage). Built with Python, FastAPI, Pydantic, SQLite FTS, pypdf, and Docker, it ensures quote traceability, not interpretive accuracy.
18 Oct 2026
NegotiazeAI
NegotiazeAI is a web based product for businesses to help increase their sales through negotiation and keep customers interested in the product. We used the Cohere API which generates more information about a product so that customers can stay engaged into buying it. The front end for this product was built through Javascript and CSS, meanwhile the backend was built through Python. Members of the team, Kudos was in charge of the product management and presentation creation. Dimitrije helped in all sorts of ways, exploring ideas, backend product development, graphics design and even helping with the presentation. Ahtesham was the genius behind the beautiful front end. He used CSS and Javascript to create a beautiful front end of the product for customer use. Kevin is a crucial member of the backend team. He helped in programming the backend and debugging it. And last but not least Maria helped us in coming up with the idea and the product recommendation system, which became a vital part of our product. Now the product is capable of generating negotiation points for a customer to keep invested in buying the product.
3 Feb 2023
