enterprise-ai on S Anand

Things I Learned - 09 Aug 2026

Sun, 09 Aug 2026 00:00:00 +0000

This week, I learned:

Kamakoti: “Entry (to the course) is relatively easy but the exit is extremely hard”. Generalizing, quality is determined by the exit criteria; loosening entry criteria is just openness / diversity. The Hindu
“Once I have a persistent system that I pay to keep thinking, learning, and acting 24/7, I think that will decisively look like AGI.” - Dan Shipper
AI has expert-level capabilities in many (increasing) tasks #ForNow. If your edge is OUTSIDE of those, use AI for other tasks you couldn’t do before, insourcing or expanding horizontally. But your edge may be short-term - so move upstream / specialize. Your competitors’ edge may be short-term, too - so plan to attack.
Analyzing Anthropic Economic Survey, it looks like people in rich countries are asking Claude for advice (explain this spreadsheet) while poor countries are asking Claude for output (build this website) #ForNow. Maybe because rich users already have tools / people that create output for them?
Humans can’t define all laws of language but LLMs have learnt them anyway. What if there are laws of nature that humans can’t understand but AI can? Actually, this is already true of black-box models (loan approvals, weather forecasts, …) where benefit/control > understanding. But as data & compute scales, this might “solve” entire fields like psychology, economics, etc. Noah Smith
In each area, there might be a limit to how much intelligence is possible/useful. For example, we’re pretty good at recognizing food and emotions - there’s not much benefit / possibility of more intelligence. But we can copy and share this intelligence - and that might help more than we think. Noah Smith
The ChatGPT Dropbox plugin can read Markdown files if you specify the path, but can only read PDF, Word, PPTX, Excel, etc. when searching. It cannot update files on Dropbox, but can add and delete. #ForNow
Given how long agents run without mistakes, verification is increasingly “drift correction”. You can’t spot it easily. Learn writing specs that EXPOSE drift. Build and test against “oracles” (verification systems). Reduce cost of error.
Permissions, in the context of multiple agents, is complex. If agent A can read my email but wants to consult agent B, can B see the email? We’d need to make permissions pretty specific, like:
- principal: “anand”
- agent: “agent-17”
- purpose: “insurance-coverage-check”
- allowed_data: [“email:read”, “dropbox/notes:read”]
- allowed_effects: [“email:send”]
- audience: [“anand”]
- expires_at: “…”
- delegation_depth: 1
After struggling to understand where to apply loop engineering, here’s my guess. If you have a metric (or something really well defined) that you want to optimize, and a single agent iteration isn’t enough, loops are a way to get there. Kaggle competitions, benchmark optimizations, etc. are examples. This means that any complex system that you can benchmark (or at least where you can robustly compare results) is loop engineerable. (This means that the ability to benchmark, and using agents to benchmark, will become a key ability.)
Ontologies, state machines, etc. can be used to create verifiable systems, e.g. nodes become states, relations are valid operations. That’s great for building verifiable systems (leading to things like LEAN). Of course, a key skill will be knowing what to put into the state, what relations to allow/disallow, what reflects reality well, how it might evolve (e.g. temporal graphs), how that might change in the future, etc.
- Having said that, this is just creating a neural network of sorts - so according to the bitter lesson, we should just toss data at an agent and have it build a graph (or not) as required. BTW, I shared this with a bunch of speakers at Data Hack Summit who were speaking about knowledge graphs. There was silence for a while. Then, gently, they all agreed.
Some people blab. Interrupting with a question is a good diversion mechanism. Some blab even after that. Exiting politely is both wise and surprisingly un-rude.
To control your mental state, breathe slowly. 5–6 times/min for five minutes (that’s longer than I thought was needed), exhaling slower than you inhale. PubMed
Once a ChatGPT conversation uses a developer plugin, #ForNow it refuses to use other plugins. So, if you need to use a GMail plugin AND a plugin you built yourself, you’d need to use the GMail plugin first, get stuff into the chat, then switch over to yours. I suspect conversations with developer plugins might not be accessible when using other plugins, too - but that’s untested.
There’s a jagged edge of AI adoption as well, not just AI capability. Several organizations limit users to weaker agents #ForNow (e.g. only Microsoft Copilot or Gemini). Many have never seen the power of Codex or Claude Code on their systems. It’s hard to convince them that AI can do much more than they think.
There’s a “data engineering” industry incentivized by structuring data. This is partly enabled by poor enterprise agent adoption #ForNow (e.g. Microsoft Copilot). The sequence works like this: “AI does not solve something with the data it’s given. Let’s structure the data. It solves it. Therefore, we need to structure data - all data.” The alternative which I believe is: agents will structure it themselves.
I noticed that when you submit a prompt on ChatGPT, it changes the URL to https://chatgpt.com/c/WEB:... and once it starts processing it on the server, changes it to https://chatgpt.com/c/... giving it the actual conversation. So, if you see a WEB: in the URL #ForNow, make sure you copy the prompt before reloading the page - because it hasn’t been saved or sent to the server.
I assumed inflammation was mostly a bio/chemical process. Looks like neural signals are involved, too, and electrical simulation can control inflammation. This leads us to a new territory: bio-electrical medicine.
The Anthropic Economic Index indicates that, on average, if you prompt Claude like an 8th grader, it responds for a 9th grader. Does that mean (a) that more sophisticated prompts get a better response, and (b) if you repeatedly meta-prompt, you increase the sophistication by about a year each iteration, and hence can get very smart prompts by just getting out of the way and with little hope of understanding the question? This might actually make sense if AI will action the result without you needing to understand.
The geometric mean is always less than or equal to the arithmetic mean. This is why a “smooth” 8% return is worth much more than a “wild” 8% return. @lumenxbt
Quantum cryptography can give us unclonable encryption, i.e. if someone copies a message midway (or you publish it), you can’t independently decrypt both. We knew how to do this in 2020. Now, ChatGPT helped “indistinguishable security”. Between 2 messages, people can’t figure out (e.g. from the length, or other attributes) which message is which. Gemini
Agents can record network requests into a HAR file and reverse-engineer an API for many websites. More efficient than browser control. dax
The Anthropic Economic Index dataset is on Hugging Face - released quarterly #ForNow. The longitudinal analysis is likely to be interesting.
BusinessCaseBench solved over 238 business cases with AI agents and they’re doing well and improving #ForNow. Not surprising. Frontier AI performance across the business disciplines
OpenAI Presence shows a pathway for deploying agents. Deploy for a specific job, with only required access to knowledge and systems, company defined policies for approval, agent periodically reviews logs & escalations and proposes updates for testing and approval.
A lot of work people are doing on ChatGPT is OUTSIDE their area of work. “… a substantial part of work-related ChatGPT use is from users expanding their role.” OpenAI

Where Enterprise AI is headed

Sat, 23 May 2026 08:24:19 +0800

A podcast host sent me eight questions. Instead of rehearsing answers in my head, I used ChatGPT with Local MCP to read 6 months of call transcripts and find the best examples:

Iteration 1: Here are questions I have been asked to answer in a podcast. Help me prepare with examples. For each question, go through my transcripts or emails and find examples relevant to the question and share (for each relevant example) a summary, how it’s relevant, and the relevant verbatim quotes from the transcript.
Iteration 2: Mention WHO said it. Emphasize the most important parts. Do a second pass. More examples. Disprove your own hypotheses with evidence to the contrary and retain what remains robust.
Iteration 3: Do a third pass. Find more real-life examples. Try and disprove yourself even harder. Share the best examples for what survives - not all. Same format.
Iteration 4: Ensure diversity of client examples. For example, in Q2, all three are the same client. Extend to add / replace examples - ideally with better ones.

Then I used Claude with examples of my writing style to summarize it in my voice.

For the first time, I’m happy to publish an AI-written blog post, because it is:

Unique to me. No one else has my transcripts, and I’m in an unusual position: enterprises keep calling me with AI problems.
More than me. Step #1 takes hours of research. Step #2 takes an hour of patience. I would not do this without AI.
Not unlike me. I would have written it slightly differently. Better in some ways, worse in others, but it’s close to my style.

Given how comfortable I am about this, I plan to be - not just an author, but also - a editor of AI generating from my content.

This article uses verbatim transcripts where possible. I’ve anonymized clients and most colleagues. I’ve annotated the post with (Anand: …) commenting with my understanding.

1. So what is Straive, and what do I do there?

I’ll let Namit explain it. He said this two weeks ago in a pitch to a European credit-insurance client:

The focus for Straive is helping its clients operationalize AI. For that, we bring two apparently distinct capabilities together… data analytics and tech development… and large-scale operations. Where we come in is bringing these together and bridging the gap.

That’s the company. We have about 8,000 people in India - Bangalore, Hyderabad, Chennai, Gurgaon, Noida, Mumbai, Kolkata, Pune. (Anand: Globally it’s probably 18K.)

My job is innovation. In the same call, I described it as:

I lead innovation at Straive. Most of my work involves playing around with Large Language Models, trying to see how they can accelerate our client work as well as deliver new kinds of solutions. That includes improving the software development life cycle.

I introduce myself as an LLM psychologist when nobody’s watching for corporate decorum. Half of my week is demos for clients. The other half is figuring out why those demos haven’t reached production yet. (More on that in question 2.)

2. Why do so many enterprise AI pilots stall?

Not for one reason. I keep a mental list of stall patterns. Three of them come up almost every week.

Pattern 1: The pilot worked, but nobody is delivering it.

In a sync with Namit a few months ago, I caught myself saying:

At a global media client, I am a little worried that the engagement keeps growing and we haven’t delivered anything yet… Right now, we’ve been given proposal after proposal after proposal… Nothing has gone to getting deployed so that someone other than our team can use it.

Namit’s reply was a single, useful sentence: “But they are not in the execution phase?” That was the gap. We had impressive demos. We had no delivery owner.

Pattern 2: The data can’t move.

For a global premium-schools group, the on-site data lead told me:

This is the data set that is at the most granular level. There are around 400,000 rows… and around 110 columns… They cannot export it… we cannot export this outside externally.

The pilot didn’t fail. The architecture failed. We had to redesign the entire engagement around the constraint: schema, profiling stats, sample rows, hypotheses, and queries flow out; raw data stays in. (Knowledge infrastructure as a workaround for missing data infrastructure. See question 4.)

Pattern 3: Teams debate frameworks instead of evals.

A private-markets investor wanted to lock the “agentic framework” by end of week. Their team was comparing LangGraph vs OpenAI Agent SDK vs Pydantic AI. I told them, more bluntly than I should have:

The technical solution may not matter too much because this is moving so fast that anything we built will anyway be outdated in not more than a year… It almost doesn’t matter which of these… the effort on the code is the least of our problems.

Pilots stall there too - not because the framework is wrong, but because the question is wrong. Without evals and acceptance criteria, no framework choice will rescue the project.

The thing that survives all three patterns: pilots prove that a model can produce a good answer once. Production proves the operating model.

3. What operational gaps stop AI from scaling?

Telemetry. Objective clarity. Repeatable loops. In that order.

Telemetry, not surveys. Our L&D lead asked me how to assess AI readiness across 19,000 employees without sounding like a particular Big Consulting firm threatening people’s promotions. I suggested:

I would go to the IT team and ask them for three things… using NetSkope, who has been accessing AI-related sites on how many unique days in the past 90 days… Regularity matters more than volume… LLM Foundry access. They have the logs for that. Third, Google Workspace tracks Gemini usage… These three give us a good company-wide proxy for AI usage.

He paused, then said it was better than a survey. You cannot scale AI adoption without knowing who is adopting it. Self-reports won’t tell you. Logs will.

Objective clarity beats agent architecture. (Anand: KISS: Keep it simple & stupid.) A teaching assistant in my IIT Madras course built an elaborate agentic workflow tool - planner agents, executor agents, sub-agents reporting to leaders. After fifteen minutes, I said:

You’ve been speaking for 15 minutes and I haven’t understood what you want. I don’t know if you understood what you want… You mentioned two objectives: learning traces and helping students learn. We should keep those as two different tools… For the learning traces, the minimal solution is a terminal command. It should authenticate them with their Google account and log all the inputs and the outputs, save it in a signed document that is tamper-proof, that we can replay.

200 lines of Python, not a multi-agent framework. (He took it well, I think.) The operational gap was: nobody had separated the two objectives, so every solution looked too complex.

Repeatable loops beat heroics. (Anand: Iterate. Compound improvement.) An internal team complained they couldn’t ship because the developers were on other work. I told them:

You can try. Don’t worry about what is not working. Just write it down. I tried this, this is working this way, this is not working in this way.

The gap wasn’t developer capacity. It was the absence of a “try, document, learn, repeat” loop that anyone could run.

4. Why does content and knowledge infrastructure matter as much as cloud?

Because the model is generic. Your business meaning is not. (Anand: Each company has their own ways of working.)

A delivery lead working at the global premium-schools client kept hitting the same wall. The bottleneck wasn’t access. It was semantics:

The real bottleneck is not access; it’s shared semantics: ‘Acceptance date,’ ‘account ID,’ ‘boarding type,’ ‘inquiry journey’ - these can mean subtly different things across systems.

That is knowledge infrastructure. Definitions. Update rules. What “acceptance date” means when a stage is updated vs appended. No model knows this until you write it down.

At the European credit-insurance pitch, we made this explicit. A senior delivery architect on our side told the client:

We create a Confluence setup, bring in everything that’s not already there on Confluence and create a comprehensive Confluence setup… That becomes the input for our agentic implementations as well. That becomes the data room from where the agents draw the knowledge to perform the actions.

The Confluence wasn’t the deliverable. It was the substrate that made every later agent deliverable possible.

On a CPG analytics product demo, the founder explained their “definition library”:

This is where we’re configuring the DNA of the agents… We call the domain definitions. We also call it the definition library… It’s not just a wrapper around ChatGPT. It’s something that’s very grounded in domain-specific definitions that avoids hallucinations, non-deterministic output.

I keep coming back to this. Cloud is where the model runs. Knowledge infrastructure is what the model knows. Skip the second, and you have a very expensive autocomplete.

5. What do India-led capability centers add?

They convert AI demos into reliable processes. That’s not a slogan. It’s the only thing that actually scales. (Anand: You need people to operate the AI machinery.)

On the European credit-insurance engagement, the client’s IT lead described his Bangalore team. Jishnu’s response was telling:

We also want to absolutely be open and also retain some of that knowledge, because as we transition, those will be critical, the knowledge that is inherent in your people and processes.

The center isn’t a labor pool. It’s a knowledge sink. Without that retention, AI workflows lose context within months.

Better example: a media-intelligence client picked us because our AI model scored higher than theirs and higher than humans. The numbers were:

Their model: 40% accuracy
Their human reviewers: 65% accuracy
Our model: 70% accuracy

But 30% of cases were still outliers. So we set up an operations team in India to handle those exceptions. AI plus humans, with the humans owning the exception path. We now have about 150 people doing similar work for a global short-video platform out of Hyderabad and Chennai. (Anand: This is a claim I heard in a pitch. I don’t have evidence. So, if it’s untrue, it’s human hallunication, not AI.)

Closer to my own work: we have a Hyderabad team that trains coding models. (Anand: Actually, we don’t. Rukesh of Deccan.ai does. This is AI hallucination.) About 100 full-time reviewers and 200-300 contractors. The full-timers don’t build models - they look at code and rate it, “I like this, this is not so good.” They’re managing reviewers, not writing code. That’s a capability center evolving from delivery to AI ops.

The thing that distinguishes India-led centers in 2026 isn’t cost. It’s the willingness to own the 30% that AI can’t handle yet.

6. Where does governance actually bite?

Three places, all real, all from the last quarter.

Compute-to-data, not data-to-cloud. (Anand: Move code, not data.) Back to the global premium-schools client. The data could not leave. So the governance pattern became:

If you could share back the output aggregated of those queries, that will be great… Get the magnitude and the P-value. Which you can dictate over a call if required.

We export queries and import aggregated results. The schema travels; the rows don’t. “No export” turned out to be a product requirement, not a blocker.

Honest impossibility. A global media client wanted us to scrub PII from 3 million user-uploaded images. Their senior engineering leader insisted on zero leaks. I did the math out loud:

For 3 million images… with… 99%, we’re talking about 30,000 images with personally identifiable information potentially slipping through.

He replied, flatly: “We have to have zero leaks. Not thousands of leaks.” I said: “Then I think I can safely say we can’t do this. This requires more technology than we have.” (Anand: When I said this, our sales teams nearly had a heart attack. So did the client, I think.)

Trustworthy AI sometimes means saying no. That was a governance decision, not a technical one.

Local execution for sensitive data. At a clinical-data conference, I used our own finance controller (a famously cautious Chennaiite) as an example. He emailed his team:

Team, please use this opportunity to install CodeX AI as per the recorded demo. This is very powerful, yesterday I tried it for two data requests and the result was fantabulous.

The reason he was comfortable: the data is not going to the model. The code is coming from the model. Codex ran the code on his machine, on the financial records, which never left his laptop. (Anand: Well, kind-of. Some data does leave, like summaries, previews, etc.)

Three governance patterns, three different problems. None of them is policy text. All of them are architecture decisions.

7. How should we measure real ROI?

Cycle time. Quality. New revenue. Risk avoided. Adoption. Not headcount.

Cycle time, hard number: on the European credit-insurance engagement, our sales lead told the client:

We brought in an AI-based approach solution to accelerate that entire mapping exercise… reduced the execution time by about 80%.

That’s the easiest ROI to defend. It was an actual XSLT and data-mapping workstream, not a demo.

Quality and effort, blended: in a workshop with our research analytics team, an analyst said a CIM (Confidential Information Memorandum) takes:

Three to four man-day effort. A man-day is equivalent to eight hours.

I did the demo live:

In approximately five minutes, Claude will come up with a pretty solid presentation. In approximately 45 minutes… ChatGPT will come up with an outrageously detailed presentation… Those three to four days will come down by 50%.

The half-day saving counts. The “five minutes vs three days” headline doesn’t, because review still takes time. Honest ROI includes the verification effort.

Revenue, not just cost: one of our innovation track leads told the team:

This was the demo that we made and that resulted into these two projects, both Sports Coverage and Trends to Clip.

That demo turned into part of a $1.15M week of deal movement. Demos that drive pipeline are an ROI line item too, even though no spreadsheet ever credits them. (Anand: This was reported in an internal sales call I was not a part of, but is true.)

The fourth measurement is adoption. If nobody uses the thing, the ROI is zero regardless of theoretical capability. Track NetSkope logs, not certificate completions. (See question 3.)

8. Where is enterprise AI going?

Three predictions, ranked by how confident I am.

Most confident: analysts stop doing research; they start managing AI researchers. I told a research analytics workshop:

Stop doing research. Your job has now transformed into somebody who has a team of 100 researchers under you… Your job is no longer managing a team; it is in fact managing a team of teams, perhaps.

The implication is real. Hiring shifts toward verification, judgment, and exception-handling. The org chart compresses but the supervisory layer grows. Accountability becomes the scarce skill.

Reasonably confident: agentic frameworks will commoditize within a year. Back to the private-markets investor sync. I told them not to obsess about LangGraph vs Pydantic AI vs OpenAI’s SDK:

Workflows are where you say, ‘do it this way.’ Agents are where you say, ‘figure it out.’… A scalable approach is to give it an agentic loop, say, ‘you figure out how to solve the problem.’

The frameworks won’t matter. The loops, tools, and evals will. Pick something boring and move on.

Less confident, but worth saying: nobody jumps straight to autonomous. Even our most ambitious proposal - the European credit-insurance one - staged it explicitly:

The Agentic AI component is something which is more prominent and starts in the Modernize phase. But it’s not there in phase one and phase two. However, AI is still there… We are going to leverage AI or LLMs for very basic functionality in a non-intrusive manner.

Assist first. Instrument. Capture knowledge. Then automate more deeply. The companies that try to skip to step four lose three quarters re-doing step one.

What survived three passes through my transcripts

I rebuilt this answer three times. Each time I tried to disprove my own pattern. Each time the same shape held:

Enterprise AI does not fail because the model is weak. It fails when we mistake a demo for delivery. When data can’t move. When teams debate frameworks before defining evals. When the team overbuilds because nobody asked what the actual objective was. When the business case never cleared cost and timeline.

The pattern that survives across clients, sectors, geographies: successful enterprise AI is operational AI. It needs telemetry, knowledge infrastructure, governance, human accountability, and a delivery model. Not slogans, not frameworks. Logs. Schemas. Confluence pages. Codex on a controller’s laptop. A 200-line terminal recorder instead of a 10,000-line agentic platform.

That’s what we sell at Straive. That’s what I’m trying to scale. Ask me again in a year.

Gemini Enterprise Business

Thu, 19 Feb 2026 10:35:09 +0800

I got an email from Google Cloud on my work account “excited to introduce you to Gemini Enterprise”.

Once I signed up, it said, “you have 30 days to try Gemini Enterprise – Business edition at no cost.” After that, it costs US $21/user/month, which I can subscribe to here.

The main differences from Gemini Pro (consumer accounts) seem to be:

Data Privacy. Google won’t read or use your data to train. (In Pro, you need to turn it off explicitly. Here, it’s the default.)
Admin Controls. Admins can turn off connectors, manage users, retention policies, etc.
Copyright Indemnification. If AI infringes copyright and you get sued, Google will find the case.

But if you’re using Gemini via your Google Workspace account (i.e. your work account already has Pro subscription), then it makes no difference - it’s all the same.

So I have no clue why Google’s mailed me to buy the same product I already have.

Features actually used in an LLM playground

Sun, 12 Jan 2025 04:28:58 +0000

At Straive, only a few people have direct access to ChatGPT and similar large language models. We use a portal, LLM Foundry to access LLMs. That makes it easier to prevent and track data leaks.

The main page is a playground to explore models and prompts. Last month, I tracked which features were used the most.

A. Attaching files was the top task. (The numbers show how many times each feature was clicked.) People usually use local files as context when working with LLMs.

3,819: Remove attachment.
1,717: Add attachment.
970: Paste a document
47: Attach from Google Drive

R. Retrieval Augmented Generation (RAG). Many people use large files as context. We added this recently and it’s become popular.

331: Enable RAG (answer from long documents)
155: Change RAG system prompt
71: Change RAG chunk size
27: Change number of RAG chunks

C. Copying output is the next most popular. Downloading is less common, maybe because people edit only parts of a file rather than a whole file.

1,243: Copy the output
883: Format output as plain text
123: Download as CSV
116: Download as DOCX

T. Templates. Many users save and reuse their own prompts as templates.

314: Save prompt as template
98: See all templates
53: Insert a template variable
18: Delete a template

J. Generate JSON for structured output is used by a few people.

238: Enable JSON output
223: Pick a JSON schema

P. Prompt optimization. Some people adjust settings to improve their prompt, or use a prompt optimizer. I’m surprised at how few people use the prompt optimizer.

238: Change temperature
207: Optimize the prompt

G. Generating code and running it via Gemini is less common, but it’s used more than I expected.

275: Generate and run code

S. Search is used a lot less than I expected. Maybe because our work involves less research and more processing.

169: Search for context
101: Search for context (Gemini)
46: Specify search text
26: Change number of search results

I left out UI actions because they do not show how people use LLMs.

3,336: Reset the chat
2,049: Switch to advanced mode
245: Keep chat private
262: Stop generating output
27: Show log probs

The main takeaway is that people mostly use LLMs on local files. We need to make this process easier. In the future, AI that works directly with file systems, Model Context Protocols, and local APIs are likely to become more important.

What does Gramener ask ChatGPT?

Sun, 14 Jan 2024 07:23:38 +0000

I looked at how Gramener uses ChatGPT Plus by evaluating 600+ chats asked over 3 months from Oct 2023 to Jan 2024.

The team asks 6 questions a day. We don't track who or how many actively use ChatGPT Plus. This also excludes personal ChatGPT accounts. Still, 6/day is low for an entire team put together.

The questions fall into 8 categories.

Category	%
Excel, data exploration & analysis	25%
Text extraction and summarization	13%
HTML, CSS, or JavaScript code	13%
Python code	13%
LLMs, AI and use cases	9%
OCR and image analysis	9%
Generate images, logos, and designs	7%
General knowledge, policy & environment	5%
Audio and translation	5%

Here are some questions from each category - to give you an idea of emergent ChatGPT Plus usage.

Excel, data exploration & analysis (25%)

Excel clean and merge. There are 2 worksheets in this excel with data, can you clean up the data and merge the data in both the sheets
Excel CO2 Data Analysis. You are an expert Data Analyst who is capable of extracting insights out of data. Analyze this sheet and let me know the findings
Excel Chi-Square Analysis Guide. how to perform chi square analysis in excel
Log Data Insights & KPIs. Looking at the columns from this excel, what kind of insights are possible, what are key KPIs to be looked at

Text extraction and summarization (13%)

Complaint Investigation Summary. The following is the summary of an internal investigation for a customer complaint. Now this internal summary is to be paraphrased (in 3-4 lines) as part of a closure
Extracting Tables from RTF. Can you write a script to extract the tables from this document
Extracting Entities from Text. [{'word1': '(P)', 'nearest_word1': 'P/N:', 'nearest_word2': '0150-25034', 'nearest_word3': 'CARTIRIDGE'}, {'word1': 'P/N:', 'nearest_word1': '(P)', 'nearest_word2': '015...
Extract PDF Font Details. Extract text formatting information from this document. Especially find font styles, families and sizes.

HTML, CSS, or JavaScript code (13%)

HTML/CSS Chart Template. Give me HTML, CSS and chart code for this design.
CSS Font Stack: Explanation. Explain this CSS font convention: Arial, Helvetica, Segoe UI, sans-serif
Checkbox Validation with JavaScript. In HTML form, I have a set of checkboxes. How do I write the form so that at least one of them being checked is mandatory?
Prevent Text Wrapping CSS. <span class="text">Chief Communications Officer</span> I need CSS such the text inside should not wrap create new line
ReactJS App with Routing. Give me developed version using ReactJS use react router for sidebar section navigation to the pages use Tailwind css for styling. Use styled components for conditional …

Python code (13%)

Python Code Documentation Guide. Can you generate documentation for a project code written in python?
Linux Commands for Python. Give me list of linux commands to work on python coding
Code explanation request. What's this code about? …
FastAPI Async Testing. Write a fastapi code and a python client to test the asynchronous nature of the fastapi package.
Streamlit App for Translation. Given the following python code, give me a simple streamlit app that takes file upload and converts that into a target language: …

An interesting sub-topic was interview question generation.

Python Decorator for Database Queries. Create one medium level question for Decorators in python Industryy usecase specific with solution

LLM, AI and use cases (9%)

LLMs for Data "What Ifs". You are an LLM Expert. Can you tell me how can we leverage LLM for implementing What IF scenarios on Data?
LLMs: Current Challenges & Concerns. what are current challenges with LLMs
LLM Applications in Marketing. Show LLM applications for the marketing function of a music company.
Gen AI usage. What industries are using Gen AI the most
Best LLMs in 2023. Search the internet for the most recent LLMs and list the best LLMs in terms of performance
Best Image Classification Models. suggest best models to tell what there in the image

OCR and image analysis (9%)

Browser history OCR. This is a screenshot of my browser history. Convert that to text. Categorize these into common topics.
Extracted C Code. This image contains C code. Extract it.
Image text extraction and annotation. Extract the text from this image and annotate the boundaries of the text
Detecting Document Image Orientation. oreientation detection of documnet image
AI Project with OpenCV & YOLO. Consider yourself as Open CV and Yolo expert and help me with AI project
Image Correction Techniques. what are the approaches we have in computer vision where my image is tilted or rotated in reverse or image is not in readable format

Generate images, logos, and designs (7%)

Google Chacha and ChatGPT Bhatija. Generate an image of Google Chacha and ChatGPT Bhatija
Regenerative Systems Group Image. Generate an Image with below context > "A group of people interested in Regenerative systems. The focus is on reusing food, energy and mental health"
Twitter Reply Icons Design. Give me three icons: icon16.png, icon48.png, icon128.png for an extension that I'm building that suggests replies to tweets
Generate flowcharts. Make a flowchart of the underlying working of a web app. Here's how it works. 1. The user uploads a document - a PDF or an image. They then select the language that …
Create Animated GIF from Photos. I have 4 photos I want to make an animated gif out of them. How can i do that?
Climate Impact Illustration. An illustration showcasing the impact of climate change on daily life, focusing on a rural setting near the coast. In the foreground, a small farm is visibly struggling, …

General knowledge, policy & environment (5%)

Design Thinking Overview. What is Design thinking
Arthashastra. What can Arthashastra teach us about modern politics?
Community Impact on Habits. Is there research to suggest the impact of community on habit building?
Focus at Age 28. What should a 28 year old focus on?
Superconductors. Explain superconductors like I'm five years old.
Climate Career: Impactful Choices. You a career counsellor at a University campus. You want to create 4 to 5 talking points for students to consider a career in Climate space.
Sustainability Division Vision. I run a software outsourced product development company. I want to start a new division that focuses on sustainability services offerings. Please draft a vision…

Audio and translation (5%)

Audio Timestamp Mapping. timestamp mapping for transcribed audio
Transcribe Lengthy Audio: Segment. Transcribe this audio file.
Traducción del MOU al Español. Translate this document to Spanish, and create a new translated document. Maintain text formatting.
Telugu Transcription into Hindi. Transcribe the following telugu text into hindi. You are supposed to transcribe, not translate. శ్రీనివాస పూజావిధానము …
GPT lacks native audio support. Does gpt support audio in audio out natively?

Open source in corporates

Tue, 04 Jun 2013 06:42:00 +0000

[This is a post that I’d published internally in InfyBlogs in Dec 2009. Time to share it.]

Last month, my first application went live.

I’ve been writing code for 20 years. Not one line of my code has been officially deployed in a corporate. (Loser…)

It’s a happy feeling. Someone defined happiness as the intersection of pleasure and meaning. Writing code is pleasurable. Others using it is meaningful.

But this post isn’t quite about that. It’s about the hoops I’ve had to jump through to make this happen.

I’ve been living in a nightmare since March 2009. That was when I decided that I’d try and get corporates to use open source.

March 2009

It began with a pitch to a VC firm. They were looking to build a content management system (CMS). Normally we’d pull together slides that say we’ll deliver the moon. This time, we put together demo based on WordPress’ CMS plugins.

The meeting went fabulously well. We said, “Here’s a demo we’ve built for you. Do you like it?” The business lead (Stuart) was drooling and declared that that’s exactly what they wanted. The IT lead (another Stuart) was happy too, but warned the business users: “Just remember: this isn’t how we do development, so don’t get your hopes up that we can deliver stuff like this :-)”

Time to make my point. I asked, “What’s your policy on open source software?”

The business lead went quiet. “I don’t know,” he finally said. Fair enough.

I turned to the IT lead. “Well, we don’t use it as a matter of policy… there are security concerns…” he said.

“Which web server do you use?”

”Oh, OK. I see what you mean. We use Apache. So on a case to case basis, we have exceptions. But generally we have security concerns.“

”Why? Do you believe open source software is more insecure than commercial software?“

He thought about it for a while. “Well… maybe. I don’t know.” We debated this a bit. Then we found the real issue: “It’s just that we don’t have control over the process. We don’t know enough about it to decide.”

A couple of weeks later, I tried pitching to a newspaper. This time, it was our sales team that raised the same question. “But… isn’t open source insecure?”

I didn’t even bother pitching any open source stuff to them. But I’d learnt my lessons:

Demo the application. Don’t talk about it.
Show it to the business first, and then tackle IT.

Aside: June 2009

In June, I got another chance at a client where we were building their new website. The very first thing I did was ask to see the Javascript. Total mess, and filled with browser-incompatible DOM requests. So I went over to their web development team.

“Look, why don’t you guys use a Javascript library? It’ll get you cross browser compatibility and compact maintainable code at the same time.”

And, to their credit, they said, “Sure. Which library?”

I showed them this and we agreed on jQuery. So, if nothing else, I’ve managed to get one open source library into a corporate.

July 2009

I was also looking at payments on the website, and our client was looking to replace their chargeback application. Since I had a week off, I built a working PCI compliant prototype on Django. (I must clarify what I mean by PCI compliant. You see, any application that stores credit card information must pass through a stringent security clearance process. I bypassed the problem by not storing the card information. I’ve realised that I’ve been building PCI compliant applications all my life – and it’s a huge benefit to let people know that.)

This time, I applied the lessons I’d learned, and demo-ed it to the business, who were thrilled. Time to tackle IT.

I started with the architecture team. Matt on the architecture team was the most approachable. So I went over, demo-ed it, and said, “Matt, this took a week to put together. It’s based on some new technologies. Are you game to try these out?”

He was. And quite enthused about it too. So we put together a proposal for the architecture review board, proposing a new technology stack: Django / Python and MySQL. As before, I showed the demo before I talked technology. I had prepared answers to all security related questions upfront (and practically memorised section 3 of the PCI guidelines.) The clincher, though, was the business case. To build it on Java, it would cost ~1,000 person days. On Django, I’d mostly done it in 5. There was no way of justifying 1,000 person days for an application that could save, at best £100,000 a year.

So they said “Go ahead, we’re fine if operations and infrastructure are fine.”

It was time to find a Django developer in Infy. I hunted for a couple of weeks but none was available. (Only 2 people that I knew knew Django in the first place.) So that effort got canned, and we were back to the 1,000 person day solution. (Which got canned too, later.)
But in the process, I’d learned my third lesson.

If you’re trying new technologies, plan on delivering it yourself.

October 2009

Another application popped up that looked like a prime candidate for introducing open source. They were using an Excel application to fraud screen orders, and wanted to make a web app out of it.

I followed the same route as before. Demo it. Show it to business first, then IT. Built it myself. I skipped Architecture, since they’d already approved the technology stack, and took it straight to Infrastructure.

“This application uses Apache as the web server, MySQL as the database, and uses PHP and Javascript for the application logic. Could we get a Linux server to host it?”

Our entire conversation lasted 30 seconds. He said, “No. We use Windows servers” (I was fine)

“… and you’ll need to chance Apache to IIS” (fine again)

“… and we don’t support PHP, so it’ll have to be Java or .NET” (I don’t know .NET or Java… but fine)

“… and we don’t support MySQL, it’ll have to be SQL Server” (fine, I guess)

“… and we don’t have DBAs available until January, so you’ll have to wait.” (definitely not good.)

So back to the drawing board on the technology stack. I needed something in Java (I know very little Java, but nothing at all in .NET) and to avoid the DBA headache, it would have to bundle in a database. I first explored key-value stores like CouchDB, Redis, etc. None of them worked on Java. The only one I found that did was Persevere, and it was a JSON data store, which fit perfectly with my plans.

By this time, I’d also learn my my fourth and most important lesson.

Don’t try to promote open source. Just deliver the application

I said, “This is a custom-built application that runs on Java. Could we get a Windows server to host it?”

The answer was “Yes”, and we had it live the next day.

PS: December 2009

The application’s deployed and running. It has about 10,000 orders fraud screened by now.
And the lessons are well learnt. So when some came over asking if there was any image resizing solution I knew off, I said: “Sure, who’s your business sponsor?” Then I went over and said, “Let me show you this ~~open source~~ application called ImageMagick. It handles aspect ratios correctly, and can crop too. Doesn’t this look professional?” Then I went over to IT and said, “~~It’s open source, so you can change it.~~ It has Java bindings, so you can integrate it into your environment. It can handle 8 3000x2400 images a second on my puny laptop. It’s used by your competitors. And I can build it for you if you like.”

I might just have my second open source entry into a corporate this year.

Comments

Satyakam Goswami 13 Jun 2013 4:24 am: Hmm interesting post , we sure should exchange our notes on these kind of experiences sometime which i guess no MBA college can teach .
Shankar V 20 Jun 2013 1:58 pm: hi Anand This is interesting. I am currently working on a development of a product and we are debating which way to go. Should we go with the standard Oracle tech stack of Oracle 12g + Weblogic + Fusion Middlware + ADF & OBIEE (for BI) and build using JDeveloper. Or should we instead go for MySQL + Hibernate + Springs FW + jQuery & Jasper/Pentaho (for BI) and develop using Eclipse? While the former is standard software from an established vendor, the latter brings down the cost of operations. Plus, we are providing product support anyway. So will that not tilt the CIO decision in our favour? Or do you think they will be apprehensive despite all this just cos it is Open Source? Will be interested to know your thoughts on this. -shankar v
Yuvi 21 Jun 2013 10:20 pm: I’m still struggling to answer a larger question of ‘why bother?’. You probably weren’t in a position where you were scared of not getting a job if you quit. Why did you stay?
Tarun 3 Jul 2013 6:52 pm: All I can say is congratulations and I am J …
Narendran 22 Jun 2013 6:37 am: Very well written and I totally agree. According to corporates, bringing in an open source software as well as open sourcing a homegrown software - both are insecure. I am yet to find a plausible explanation. Apart from the problem of reinventing a wheel, this greatly limits a software developer’s versatility (unless he gets his personal time to play around with tech). Eventually, he ends up using the one big hammer to nail down all petty problems (irrespective of the type, complexity and scope of the problem). But ultimately, as yuvi says, why bother about all these when you have a good command over technologies. Get to a place that appreciates and supports you instead of fighting an endless war.
Naveen 6 Sep 2013 7:52 am: Open source is viewed by lot of corporates as “free” , I realize that anything given free is not of any value ….Nice once Anand, I can echo with lot many questions I asked you during our initial days of Gramener….you would have written me off as one other guy with same questions :-)
Ashish 5 Jun 2013 9:09 am: Hi Anand, I am no technie but I find your dabbling (as I see) and outcomes of various nice tools interesting and been following you for couple of years now - started when I found your search music by name application. So much for introduction. Recently, I was looking for list of Hindi and English movie names to build an automated application (in Excel) to randomly pick a movie. Basically, motivation was to play Dumb Charade when there are not enough members so that we don’t lose out one person who is supposed to give movie name. I couldn’t find any easy to download list though BollywoodHungama and IMDB have them in format that you can probably scrape and build whole thing in less than an hour. So would you help me? It could be interesting tool for you too! If not, I just need dump of all movie names in txt line by line for my purpose. To filter out list since it is going to be too big and for me to process in Excel, we may put IMDB rating > 4 restriction or some such.