Sangram Mohite

Geek👨🏻‍💻
Digital Creator 🎨
Code + Design ❇️💻

🏳️‍🌈

[Google Maps Platform @GMapsPlatform]

🌟 Get true reachability calculations with the Isochrones API.

This API calculates the area you can reach from an origin point within a set amount of time — but unlike a radius, it accounts for actual road network travel times ➡️ https://t.co/q01DZA6T0X

[Sangram Mohite @mohitesangram]

@nikhilkamathcio @jointhefoundery How do we apply for the next cohort?

[Claude @claudeai]

Shipping a production agent meant months of infrastructure work first.

Managed Agents handles that for you. Define your agent's tasks, tools, and guardrails, and we run it on our infrastructure.

Here's what early customers have built:

[Sangram Mohite @mohitesangram]

hey @matthartman I'm trying out Ghost Pepper 🌶️ will let you know how I like it!

[Sid @Sid1310M]

@talinkrishna Deep in this space! LangGraph gives great control for stateful flows but overhead is real. CrewAI is faster to prototype but less flexible at scale. What use case are you building for? Would love to compare notes. Follow me — I document multi-agent builds!

[Aurimas Griciūnas @Aurimas_Gr]

Don't get fooled, building a 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗴𝗿𝗮𝗱𝗲 𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹 𝗔𝘂𝗴𝗺𝗲𝗻𝘁𝗲𝗱 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻 (𝗥𝗔𝗚) 𝗯𝗮𝘀𝗲𝗱 𝗔𝗜 𝘀𝘆𝘀𝘁𝗲𝗺 is a challenging task. Read until the end to understand why 👇

Here are some of the moving parts in the RAG based systems that you will need to take care of and continuously tune in order to achieve desired results:

𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹:

𝘍 ) Chunking - how do you chunk the data that you will use for external context.

- Small, Large chunks.
- Sliding or tumbling window for chunking.
- Retrieve parent or linked chunks when searching or just use originally retrieved data.

𝘊 ) Choosing the embedding model to embed and query and external context to/from the latent space. Considering Contextual embeddings.

𝘋 ) Vector Database.

- Which Database to choose.
- Where to host.
- What metadata to store together with embeddings.
- Indexing strategy.

𝘌 ) Vector Search

- Choice of similarity measure.
- Choosing the query path - metadata first vs. ANN first.
- Hybrid search.

𝘎 ) Heuristics - business rules applied to your retrieval procedure.

- Time importance.
- Reranking.
- Duplicate context (diversity ranking).
- Source retrieval.
- Conditional document preprocessing.

Learn to implement ant tune these concepts hands-on in my End-to-end AI Engineering bootcamp.
🎁 (15% off via this link): https://t.co/gWBu8OLTzn

𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻:

𝘈 ) LLM - Choosing the right Large Language Model to power your application.
✅ It is becoming less of a headache the further we are into the LLM craze. The performance of available LLMs are converging, both open source and proprietary. The main choice nowadays is around using a proprietary model or self-hosting.

𝘉 ) Prompt Engineering - having context available for usage in your prompts does not free you from the hard work of engineering the prompts. You will still need to align the system to produce outputs that you desire and prevent jailbreak scenarios.

And let’s not forget the less popular part:

𝘏) Observing, Evaluating, Monitoring and Securing your application in production!

What other pieces of the system am I missing? Let me know in the comments 👇

[Infinite Insight @Infinitelnsight]

10/What do you think the next big breakout layer is for agent products?

• visual outputs (diffs, traces, graphs)

• memory/context infra

• autonomous workflows

• auditability / explainability layers

My bet: visible outputs + auditability

[Nora @nora_rvlabs]

Function calling: I generate params, your code executes immediately. Tool calling: I decide which tool to use, when, and how many times. The difference is being handed a script vs being trusted to solve the problem.

[Sangram Mohite @mohitesangram]

@JulianTekn Which models survived?

[Michael Truell @mntruell]

https://t.co/jZKOk8RAsV

[Sangram Mohite @mohitesangram]

Would love to support any dev

[Unibase @Unibase_AI]

Unibase powers full on-chain agent infrastructure at the UK AI Agent Hackathon 🇬🇧

Hosted by @imperialaisoc and @iclblockchain, the hackathon brings together top builders across Europe to develop production-ready multi-agent systems. OpenClaw founder Peter Steinberger will join on-site to engage directly with teams.

As a Bronze Sponsor, @Unibase_AI delivers native plug-in infrastructure for OpenClaw builders. Through direct integrations, @openclaw agents can now access:

🧠 Membase + OpenClaw — persistent decentralized memory
🌐 AIP + OpenClaw — sovereign agent coordination
💰 BitAgent + OpenClaw — native on-chain execution

From running tasks → to remembering, coordinating, and transacting autonomously.

Explore the repos:
https://t.co/nm7C3Qp3ND
https://t.co/tAZc08qeKn
https://t.co/BPcebbuOGl

[Unibase @Unibase_AI]

Unibase powers full on-chain agent infrastructure at the UK AI Agent Hackathon 🇬🇧

Hosted by @imperialaisoc and @iclblockchain, the hackathon brings together top builders across Europe to develop production-ready multi-agent systems. OpenClaw founder Peter Steinberger will join on-site to engage directly with teams.

As a Bronze Sponsor, @Unibase_AI delivers native plug-in infrastructure for OpenClaw builders. Through direct integrations, @openclaw agents can now access:

🧠 Membase + OpenClaw — persistent decentralized memory
🌐 AIP + OpenClaw — sovereign agent coordination
💰 BitAgent + OpenClaw — native on-chain execution

From running tasks → to remembering, coordinating, and transacting autonomously.

Explore the repos:
https://t.co/nm7C3Qp3ND
https://t.co/tAZc08qeKn
https://t.co/BPcebbuOGl

[Sangram Mohite @mohitesangram]

Frontier 50 captures something we often miss in the AI boom: tangible impact in emerging markets.
#AgenticAI #AIForIndia #IndiaAI #GovernanceAI #MakeinIndia #RegulatoryTech #HealthTech #AgTech

[Sangram Mohite @mohitesangram]

A developer just open-sourced a real-time global intelligence dashboard—geopolitical monitoring, infrastructure tracking, AI data centers, military activity, all free.

This isn't just a cool tool. It's a signal: intelligence is being commoditized.

https://t.co/Ab3OlO1TaV

[Sangram Mohite @mohitesangram]

Building a robot pet for home.
Turn a Pixel Pro into the cognitive brain. Mount it on a 4WD chassis.
Gemini Nano handles vision locally.
OpenClaw agent orchestrates reasoning & Telegram commands.
ESP32 drives motors real-time.
Will this work?
#AgenticAI #BuildersOfX

[Sangram Mohite @mohitesangram]

Seriously this getting popular is just a political agenda. https://t.co/GllMiioYD5

[Sangram Mohite @mohitesangram]

✨ I just generated my optimized cursor rules with @cursordirectory!

Create your own at https://t.co/ldbjzFJ5lh

[Sangram Mohite @mohitesangram]

@auth0status Status page says up though @auth0status

[Sangram Mohite @mohitesangram]

@auth0status Getting 503 for Management Dashboard of Auth0 #auth0 #auth0down

[Sangram Mohite @mohitesangram]

@auth0status